Individual and Social Factors in Crying Appraisal and Recovery: An Exploration into the Functions of Crying

Leah Sharman

Bachelor (with Honours) of Psychological Science, University of Queensland

A thesis submitted for the degree of Doctor of Philosophy at The University of Queensland in 2019 School of Psychology ii

Abstract We are currently thought to be the only animal that cry emotional tears throughout the lifespan, with crying in humans apparent in infants, children, and adults. Crying in adulthood has been theorised by a number of researchers in the field to have an intrapersonal, self- soothing, and cathartic effect on the crier. Others have argued that crying has a social (interpersonal) function by attracting the attention of people who may provide us with social support and resources. The research reported in this thesis explores our understanding of adult crying and how it functions in individual and social settings to either help or hinder emotional recovery according to a Multifunctional Model of Crying. This thesis is comprised of six chapters, with the first chapter providing a comprehensive review of previous research on crying and the second summarising the various methodologies used in crying research, while also capturing the various difficulties in eliciting and measuring crying. The third chapter reports on a psychophysiological investigation into perspectives on crying that see crying as aiding physical and emotional recovery, as well a first effort to understand if crying acts as a barrier to immediate future stressors. Compared to those who did not cry to the stimuli and those who were exposed to neutral videos, I predicted that people who cried would (a) be able to withstand a stressful task for longer; (b) show lower hormonal stress levels following crying and exposure to the stressor; and (c) have faster recovery (i.e., return to baseline levels of affect). Heart rate and respiration were continuously recorded while salivary samples for cortisol were taken at four separate time points during testing. Analyses revealed no differences between the three groups in time withstanding the stressor or cortisol changes. Respiration rate, however, increased in the neutral group and non-criers while watching the videos, with criers’ respiration remaining stable. Furthermore, heart rate was found to significantly decrease heart rate just before crying, with heart rate returning to baseline during crying. These results suggest that rather than a physical barrier to stress or pain specifically, crying may engage the body in attempts to maintain homeostasis, perhaps both consciously through self-soothing via purposeful breathing, and unconsciously through regulation of heart rate. The next chapter reports on the development and psychometric validation of the Beliefs about Crying Scale (BACS), a new measure assessing beliefs about whether crying leads to positive or negative emotional outcomes in individual and interpersonal contexts. This yielded three subscales: Helpful Beliefs, Unhelpful-Individual Beliefs, and Unhelpful- Social Beliefs. These subscales showed differential relationships with measures of personality traits, crying proneness, emotion regulation and expressivity, and emotional iii identification (alexithymia). Overall, the BACS provides a nuanced understanding of beliefs about crying in different contexts and helps to explain why crying behaviour may not always represent positive emotion regulation for the crier. Chapter 5 reports a large, international study of the role of gender, culture, and social reactions to crying. Data were collected from participants in Australia, Croatia, the Netherlands, Thailand, and the United Kingdom. In understanding the role of gender, the results suggest that across cultures a person’s gender and their adherence to gender roles is highly related to behavioural crying responses, but not related to evaluations of crying. How a person evaluates crying, instead, appears to be highly related to our beliefs about the helpfulness of crying, with no apparent relationship to gender. Crying in social contexts showed that a person is more likely both to cry and to feel that they received help around a person that they know, compared to a stranger. Furthermore, closeness to persons present during crying did not affect whether help was provided. Importantly, when a crier said that they were helped, they tended to report feeling better following crying than those who did not receive help from those around them. The final chapter concludes the thesis by discussing its findings from and their relationship with elements of a Multifunctional Model of Crying developed for this thesis. Overall, this thesis finds that individual differences, social norms, biological factors, and the presence of others all play a role in how people evaluate and understand their crying experiences. iv

Declaration by Author

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, financial support and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my higher degree by research candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis and have sought permission from co- authors for any jointly authored works included in the thesis.

v

Publications Included in This Thesis 1. Sharman, L. S., Dingle. G. A., Vingerhoets, A. J. J. M., Vanman, E. J. (2019). Using crying to cope with emotional and physical pain: Physiological responses to stress following emotional tears. Emotion. https://doi.org/10.1037/emo0000633 2. Sharman, L. S., Dingle. G. A., & Vanman, E. J. (2019). Does Crying Help? Development of the Beliefs About Crying Scale (BACS). Cognition and Emotion, 33, 722–736. https://doi.org/10.1080/02699931.2018.1488243 3. Sharman, L. S., Dingle. G. A., Baker, M., Fischer, A., Gračinan, A., Kardum, I., Manley, H., Manokara, K., Pattara-angkoon, S., Vingerhoets, A. J. J. M., Vanman, E. J. (2019). The relationship of gender roles and beliefs to crying in an international sample. https://doi.org/10.3389/fpsyg.2019.02288 vi

Other Publications During Candidature

Peer-reviewed Journal Publications 1. Williams, E., Dingle, G., Calligeros, R., Sharman, L.S., & Jetten, J. (2019). Enhancing mental health recovery by joining arts-based groups: a role for the social cure approach. Arts and Health. https://doi.org/10.1080/17533015.2019.1624584

2. Price, E.*, Sharman, L.S.*, Douglas, H., Sheeran, N., & Dingle, G.A. (2019). Experiences of reproductive coercion in queensland women. Journal of Interpersonal Violence. https://doi.org/10.1177/0886260519846851 *co-first authors

3. Sharman, L.S., Douglas, H., Price, E., Sheeran, N., & Dingle, G.A. (2018). Associations between unintended pregnancy, domestic violence, and sexual assault in a population of Queensland women. Psychiatry, Psychology and Law. Accepted without revision, July 2018. https://doi.org/10.1080/13218719.2018.1510347

4. Loxton, N. J., Mitchell, R., Dingle, G. A., & Sharman, L. S. (2016). How to tame your BAS: Reward sensitivity and music involvement. Personality and Individual Differences, 97, 35-39. https://doi.org/10.1016/j.paid.2016.03.018

Book Chapters

5. Dingle, G., Sharman, L., & Larwood, J. (2019). Young people’s uses of music for emotional immersion. Chapter for Oxford Handbook of Music and Adolescence. Oxford: Oxford University Press.

Government and Public Reports

6. Utting, S., Douglas, H., Sheeran, N., Dingle, G. A., Bell, J., & Sharman, L. S. (2018). The cost of abortion access in Queensland: Data from the Children by Choice financial assistance program for disadvantaged women 2015-2017. Brisbane, Australia: Children by Choice.

7. Sharman, L.S., Douglas, H., Price, E., Sheeran, N., & Dingle, G.A. (2018). Experiences of QLD women seeking support for unplanned pregnancy. Report submitted to Queensland Parliament.

8. Dingle, G., Williams, E., Sharman, L., & Jetten, J. (2016). School of Hard Knocks QLD – Final Evaluation Report. School of Psychology, The University of Queensland.

9. Dingle, G., Williams, E., & Sharman, L. (2015) Interim report of the School of Hard Knocks QLD Evaluation. School of Psychology, The University of Queensland.

Online Articles

10. Sharman. L.S. (2018, November). No, crying doesn't release toxins, though it might make you feel better... if that's what you believe. The Conversation. https://theconversation.com/no-crying-doesnt-release-toxins-though-it-might-make-you- feel-better-if-thats-what-you-believe-106860

11. Sharman. L.S., & Dingle. G. (2015, August). Getting a bad rap: Why problem music isn’t vii

really a problem. The Conversation. https://theconversation.com/getting-a-bad-rap-why- problem-music-isnt-really-a-problem-46533

Conference Abstracts

12. Stathis, A., Williams, E., Sharman, L.S, Nelson, N., & Dingle, G.A. (2018, May). The use of music for emotion regulation across the adult lifespan. Poster presented at the Centre for Health Outcomes Innovation and Clinical Education conference, Brisbane, Australia.

13. Sharman, L.S., Dingle, G., Vingerhoets, A.J.J.M., & Vanman, E. (2018, May). Does Crying Really Make You Feel Better? A Laboratory Investigation of the Intrapersonal Function of Tears. Presented at the inaugural Institute for Social Neuroscience and Centre for Psychology and Evolution Colloquium, Brisbane, Australia.

14. Sharman, L.S., Dingle, G., & Vanman, E. (2018, April). Coping Through Crying: A Laboratory Investigation of the Intrapersonal Function of Tears. Presented at the Consortium of European Research on Emotion, Glasgow, Scotland. Symposium organiser/convener Emotion for “The Intrapersonal and Interpersonal Functions of Crying”

15. Williams, E., Dingle, G., Calligeros, R., Sharman, L.S., & Jetten, J. (2017, December). A thematic analysis of the benefits and challenges of participating in arts-based groups for mental health recovery. Paper presented to the Australian Music and Psychology Society conference, Brisbane, Australia.

16. Williams, E., Dingle, G., & Sharman, L.S. (2017, December). The use of music for emotion regulation across the adult lifespan. Paper presented to the Australian Music and Psychology Society conference, Brisbane, Australia.

17. Sharman, L.S., Vanman, E. J., & Dingle, G. A. (2017, November). Coping Through Crying: A Laboratory Investigation of the Intrapersonal Function of Tears. Poster presented at the University of Queensland Psychology HDR day, Brisbane, Australia.

18. Sharman, L.S., Vanman, E. J., & Dingle, G. A. (2017, September). Coping Through Crying: A Laboratory Investigation of the Intrapersonal Function of Tears. Poster presented at the Centre for Health Outcomes Innovation and Clinical Education conference, Brisbane, Australia.

19. Sharman, L.S, Dingle, G., & Vanman, E. (2017, July). Why Do We Cry? A Laboratory Investigation of the Intrapersonal Function of Tears. Presented at the symposium on emotional crying at the International Society for Research on Emotions, St Louis, USA.

20. Sharman, L. S., Dingle, G., & Vanman E. (2016, November). Coping Through crying. Poster presented at the Australasian Cognitive Neuroscience Society Conference.

21. Sharman, L.S., Vanman, E.J., & Scambler, J. (2015, December) The Social Function of Tears in Crying. Presented at the 25th Annual Australasian Society for Psychophysiology Conference, Sydney, Australia.

viii

Contributions by others to the thesis Reneé Calligeros and Hannibal Thai contributed to data collection of the psychophysiology study. Dr Ad Vingerhoets contributed to the design of the psychophysiology study and provided a review of the manuscript. Please see Chapter 3 for more details.

Data collection, translation, and/or review of the manuscript for the cross-cultural study was contributed by Dr Asmir Gračinan, Dr Ad Vingerhoets, Dr Agneta Fischer, Mr Kunalan Manokara, Dr Harrison Manley, Ms Pattara-angkoon, and Mr Marc Baker. See Chapter 5 for more details.

Associate Professors Eric Vanman and Genevieve Dingle provided input into the theoretical aspects and design of each of the studies within this thesis. Dr Vanman and Dr Dingle also reviewed and commented on this thesis and the manuscripts within it.

ix

Statement of parts of the thesis submitted to qualify for the award of another degree No works submitted towards another degree have been included in this thesis.

Research Involving Human or Animal Subjects All research within this thesis involved human subjects and received prior ethical review approval by an independent review committee within the UQ School of Psychology, and the UQ human research ethics committee. Ethics approval information can be found in Appendix D. x

Acknowledgements

First and foremost, I would like to thank both of my advisors, Eric Vanman and Genevieve Dingle for providing me with unwavering support and being my biggest champions throughout the 7 years I have known both of you. I absolutely could not have done this without you, and this would have been a different rollercoaster if we had never crossed paths. Thank-you for the endless reading of drafts and feedback, thank-you for the BBQ’s, Champaign/wine/Pimm’s celebrations, and thank-you for giving me confidence in every aspect of my work.

To my partner, Jarred, thanks, but no thanks for having us renovate a unit for over a year in the middle of this PhD. Your determined attitude has helped me so much throughout the highs and lows of this experience and I can’t thank you enough for being there for me - always. Thanks for all the lunch-time oranges, strange dinner innovations, and your excellent cat-foster-dad abilities. Without you I would have forgotten that there is more to life than this PhD but I’m sure you’re just as glad as me that it’s now over.

Thanks to all the incredible people I have met and lost in my time volunteering at the aged care home in Bardon. You have all been a shining light every week, just to be in your presence has been a joy. I also have to acknowledge all the animals Jarred and I have fostered over this period, some of which have made it to the care home for cuddles. Without you all I would have been a much less put-together human being, your ability to love and trust us after experiencing significant trauma is treasured and will never be forgotten.

Thank-you to all of my friends and family over the last few years. Sorry about all the sporadic visits and large stretches of time where I was absent. I have appreciated all of the gin nights, drag brunches, dinners, gigs, cuddles with my nephews, and just having you all there when I needed you - and when I didn’t.

Finally, thank-you to me for actually finishing this god-forsaken thing.

xi

Financial support This research was supported by an Australian Government Research Training Program Scholarship

Keywords crying, tears, physiology, beliefs, scale, emotion regulation, gender roles, social, cross- cultural

xii

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 170199, Psychology not elsewhere classified, 60% ANZSRC code: 170101, Biological Psychology, 20% ANZSRC code: 170105, Gender Psychology, 20%

Fields of Research (FoR) Classification FoR code: 1701, Psychology, 100% xiii

Table of Contents

ABSTRACT ...... II

DECLARATION BY AUTHOR ...... IV

ACKNOWLEDGEMENTS ...... X

LIST OF TABLES...... XVI

LIST OF FIGURES ...... XVII

LIST OF ABBREVIATIONS ...... XVIII

CHAPTER 1: THE HISTORY AND THEORIES OF CRYING AND ITS FUNCTIONS ...... 19

A DEFINITION ...... 20

CRYING ANTECEDENTS...... 21

UNIVERSALITY AND CULTURAL VARIATION ...... 22

BOYS DON’T CRY...... 24

THE UTILITY OF CRYING ...... 25 Individual Functioning ...... 26 Social Functioning ...... 29

A MULTIFUNCTIONAL PERSPECTIVE ...... 32 Testing the Multi-functional Model: An Overview of the Thesis ...... 34

CHAPTER 2: INDUCING AND MEASURING CRYING: A METHODOLOGICAL REVIEW ...... 37

WHO CRIES AND WHY ...... 38

RESEARCH METHODS ...... 40

MEASURING CRYING ...... 42 Retrospective Self-report ...... 42 In the Laboratory ...... 44 Social Context ...... 45

INDUCING CRYING ...... 48

CONCLUSION ...... 51

CHAPTER 3: USING CRYING TO COPE: PHYSIOLOGICAL RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS ...... 53

ABSTRACT ...... 54 How does crying make us feel? ...... 56 Crying Theories ...... 60 Current study ...... 61

METHOD ...... 62 Participants ...... 62 Stimuli ...... 62 Measures ...... 64 xiv

Procedure ...... 65

RESULTS ...... 67 Data processing ...... 67 Self-report ratings of negative affect ...... 68 Self-report ratings of negative affect ...... 70 Cortisol ...... 71 Cold Pressor Test...... 72 Physiological Measures ...... 73 Exploratory Results ...... 76 Sensitivity Checks for CPT Endurance and Physiological Measures...... 77

DISCUSSION...... 78 Heart Rate and Respiration ...... 80 Crying and Emotion ...... 81 Conclusion ...... 83

CHAPTER 4: DOES CRYING HELP? DEVELOPMENT OF THE BELIEFS ABOUT CRYING SCALE ...... 85

ABSTRACT ...... 86 This study ...... 91

CONSTRUCTION OF THE BACS ...... 92 Item Generation ...... 92 Item Selection ...... 93

EXPLORATORY FACTOR ANALYSIS...... 94 Participants ...... 94 Statistical Approach and Results ...... 94

CONFIRMATORY FACTOR ANALYSIS AND DISCRIMINANT AND CONVERGENT VALIDITY ...... 98 Participants ...... 98 Criterion Measures ...... 98 Results ...... 101

GENERAL DISCUSSION ...... 105 Psychometric Properties of the BACS ...... 105 Gender ...... 106 Crying, Emotion Regulation, and Emotion Expression ...... 107 Crying and Personality ...... 108 The Question of Causality ...... 109 Conclusion ...... 110

CHAPTER 5: GENDER ROLES, CRYING BELIEFS, AND THE EFFECT OF SOCIAL PRESENCE ON CRYING ACROSS CULTURES ...... 112

ABSTRACT ...... 113 The Role of Culture ...... 114 xv

Gender and Emotion Expression Norms ...... 116 Crying Beliefs ...... 117

CRYING AND THE SOCIAL CONTEXT ...... 120

METHOD ...... 123 Exclusion Criteria ...... 123 Participants ...... 123 Measures ...... 124 Procedure ...... 127

RESULTS ...... 128 Crying Relationships and Mediation ...... 128 Social Context Effects...... 131

DISCUSSION...... 134 Gender Role Endorsement ...... 134 Crying and Social Help ...... 136 Limitations ...... 138 Conclusion ...... 139

CHAPTER 6: DISCUSSION AND CONCLUSION ...... 142

THE MULTIFUNCTIONAL MODEL OF CRYING ...... 143 Biological Factors ...... 144 Individual Differences ...... 146

THE FUTURE OF THE MODEL ...... 148

CONCLUSION ...... 151

REFERENCES 152 APPENDICES 172

APPENDIX A ...... 172

APPENDIX B ...... 182

APPENDIX C ...... 183

APPENDIX D ...... 185

xvi

List of Tables

TABLE 2.1: PROPORTION OF CRIERS TO SHORT VIDEOS AND FILM EXCERPTS 50

TABLE 3.1: SUMMARY OF CRY-INDUCTION LABORATORY STUDIES AND OVERALL NUMBERS OF CRIERS ANALYSED IN EACH STUDY 59

TABLE 3.2: MEANS (SDS) AND SAMPLE SIZES FOR NEGATIVE AFFECT RATINGS, HEART RATE, RESPIRATION, AND CORTISOL ACROSS EACH GROUP AT EACH TIME POINT IN ANALYSES PERFORMED 69

TABLE 4.1: PSYCHOLOGICAL VARIABLES POTENTIALLY RELATED TO BELIEFS ABOUT CRYING 93

TABLE 4.2: FACTOR LOADINGS OF THE BELIEFS ABOUT CRYING SCALE 97

TABLE 4.3: SCALE MEANS, STANDARD DEVIATIONS, INTER-CORRELATIONS, AND CRONBACH’S ALPHA FOR INDIVIDUAL SCALE RELIABILITY FOR THE EFA AND CFA SAMPLES 103

TABLE 4.4: CORRELATIONS BETWEEN BACS AND OTHER SELF-REPORT MEASURES FOR THE THIRD SAMPLE 104

TABLE 5.1: PARTICIPANT DEMOGRAPHICS FOR EACH COUNTRY, INCLUDING NUMBER OF PARTICIPANTS BY GENDER, MEAN (SD) AGE IN YEARS, AND THE PERCENTAGE OF PARTICIPANTS WHO INDICATED THEY WERE BORN AND RAISED IN THAT COUNTRY 124

TABLE 5.2: MEAN (SD) CRYING BEHAVIOURS AND EVALUATIONS SEPARATED BY COUNTRY AND GENDER 130

TABLE 5.3: CORRELATIONS BETWEEN SELF-REPORTED GENDER, GENDER ROLES, BELIEFS THAT CRYING IS HELPFUL (BACSHELPFUL), CRYING INTENSITY, AND EMOTION CHANGE FOLLOWING CRYING 130

TABLE 5.4: RESULTS OVERALL AND ACROSS COUNTRIES TESTING IF THE RELATIONSHIP BETWEEN GENDER ROLE ENDORSEMENT AND CRYING INTENSITY IS MEDIATED BY BELIEFS THAT CRYING IS HELPFUL 132

xvii

List of Figures

FIGURE 1.1. A NEW MULTIFUNCTIONAL MODEL OF CRYING 34

FIGURE 3.1. SELF-REPORT RATINGS FOR NEGATIVE AFFECT FOR EACH GROUP AT BASELINE, AFTER VIDEO WATCHING, FOLLOWING THE COLD PRESSOR TEST (CPT), AND AT THE END OF THE STUDY. HIGHER SCORES RELATE TO WORSE MOOD. 69

FIGURE 3.2. MEAN CORTISOL LEVELS FOR EACH GROUP MEASURED AT BASELINE, FOLLOWING VIDEO WATCHING, AFTER THE COLD PRESSOR TEST (CPT), AND AT THE END OF THE STUDY. 73

FIGURE 3.3. MEAN HEART AND RESPIRATION RATE MEASURED AT 4 TIME-POINTS: AT BASELINE, AN AVERAGE OF THE SADDEST POINTS OF FIVE VIDEOS, THROUGHOUT THE COLD PRESSOR TEST (CPT), AND DURING THE FINAL MINUTE OF THE EXPERIMENT. 75

FIGURE 3.4. MEAN HEART RATE FOR CRIERS MEASURED THREE TIME-POINTS: BASELINE, PRE- CRYING, AND INITIAL CRYING PERIODS. 78

FIGURE 5.1. MODEL OF PREDICTED PATHWAYS BETWEEN GENDER ROLES, BELIEFS ABOUT CRYING, AND CRYING INTENSITY. 120

FIGURE 5.2. PLOT OF EMOTIONAL CHANGE RATINGS WHETHER HELP WAS PROVIDED WHEN CRYING IN A SOCIAL CONTEXT COMPARED BY COUNTRY. 133

FIGURE 6.1. UPDATED MULTIFUNCTIONAL MODEL OF CRYING 144

xviii

List of Abbreviations

AUCG Area under the curve with respect to ground

AUCI Area under the curve with respect to increase

BACS Beliefs About Crying Scale

BFI Big Five Inventory

BPM Beats per minute

CFA Confirmatory Factor Analysis

CFI The comparative fit index

CPS Crying Proneness Scale

CPT Cold Pressor Test

DERS Difficulties in Emotion Regulation Scale – short form

EES Emotion Expressivity Scale

EFA Exploratory Factor Analysis

HR Heart rate

IERQ Interpersonal Emotion Regulation Questionnaire

IRI Interpersonal Reactivity Index

K10 The Kessler Psychological Distress Scale

PANAS The Brief Positive and Negative Affect Scale

RMSEA The root mean squared error of approximation

SRMR The standardized root mean residual

TAS-20 Toronto Alexithymia Scale

TLI Non-normative fit index (Tucker-Lewis Index)

TOST Two one-sided t-tests THE HISTORY AND THEORIES OF CRYING 19

CHAPTER 1: The History and Theories of Crying and its Functions

“There is a certain pleasure in weeping; grief finds in tears both a satisfaction and a

cure.”

Ovid, 43 BC – 17/18 AD

CHAPTER 1 20

For centuries, humans have wondered why they cry. These considerations have led to questions such as where tears come from (e.g., the heart or brain), why crying seems to bring relief, and whether crying is functional, or merely an accidental by-product of our evolution

(Balsters, Krahmer, Swerts, & Vingerhoets, 2013; Horstmanshoff, 2014). Historically, tears have formed a central role in our creation and interpretation of art, religion, mythology, and literature; however, in many parts of the world tears continue to serve as part of cultural and spiritual healing (Vingerhoets, 2013). Social and cultural acceptance of crying as ‘good’ or

‘bad’ for both men and women has fluctuated over time, and has been viewed as both a mark of strength or as a weakness (Fischer, Rodriguez Mosquera, van Vianen, & Manstead, 2004;

Lutz, 1999). Presently, crying is considered intrinsically entwined with feelings of sadness and distress, and yet is also overwhelmingly viewed as a useful way to express emotions (van

Hemert, van de Vijver, & Vingerhoets, 2011). Though for adults in particular, crying is more than just an expression of distress, it is a unique form of emotional expression that can accompany a variety of experiences and feelings, including that of empathy, anger, joy, grief, and beyond. It is a complex intersection of psychological, cognitive, and social processes

(Vingerhoets, Boelhouwe, van Tilburg, & van Heck, 2001; Vingerhoets, Cornelius, van Heck,

& Becht, 2000). So, despite the social and cultural significance attributed to crying historically and globally, until recently, questions surrounding adult crying have received little attention in scientific research.

A Definition

Emotional crying is a complex secretomotor phenomenon, which involve tears being shed from the lacrimal apparatus (the physiological system containing the structures for tear production) that occur without the irritation of ocular structures (Patel, 1993). It is often accompanied by muscular contractions around the eyes and face (that form facial expressions often associated with distress), vocalisations, and sobbing (convulsive in/exhalation of air). THE HISTORY AND THEORIES OF CRYING 21

While others have referred to this emotional crying as ‘weeping’ to differentiate the shedding of tears from crying as a vocalisation, in this thesis I employ the word crying to emphasize that crying often incorporates tearfulness and vocalisations, and further, to highlight the similarity of tearful emotional crying in adults to infant distress calls. Importantly, this thesis advocates the view that crying is communicated on a spectrum, from tears resting in eyes through to tears rolling down a face, expressed with muscular contractions, vocalisations, and sobbing. Accordingly, in this thesis I consider (emotional) crying to be tears which are shed when there is no physical irritation to the eye, in particular when caused by a negative antecedent, such as sadness or helplessness, and which occurs with or without facial changes, vocalisations, and/or sobbing.

Crying Antecedents

Crying is a universal form of human emotional expression (and believed for humans alone), which is expressed through most of our range of emotion, from joy to anger or sadness. There is considerable variation in the reasons we cry, with our expression of crying initiated by any number of possible situational and emotional antecedents. We cry because of laughter, pride, and relief, as well as sadness, anger, frustration, and anxiety. We also cry because of aesthetic experiences, such as a feeling of being moved or touched by a beautiful image, music, opera, or film.

The literature in this area has approached the question of ‘why do people cry?’ in many different ways, with the answer generally agreed, for lay people and academics, that crying is primarily caused by sadness. Indeed, much of the research and theoretical perspectives in this area have considered crying as synonymous with feelings of sadness.

Among the reasons most often cited to explain why people cry are: loss, romantic break- downs, and sad films and television; yet, there also frequent instances where people indicate they cry from happiness because of music, weddings, and reunions (see Vingerhoets, 2013). CHAPTER 1 22

However, reports of most recent crying episodes do not always match these more global (and arguably rare) antecedents, with people often citing small and idiosyncratic reasons for crying.

For example, conflicts with others, minor failures, criticism, or rejection are among some of the more common situations that cause crying.

Though very little work has been published on crying antecedents, other than sadness, it has been suggested that crying is primarily caused by feelings of helplessness or powerlessness, and indicates an inability to cope with a situation (Frijda, 1986). Miceli and

Castelfranchi (2003) also proposed that all crying has a common core intertwined with helplessness. That is, crying for any emotional or aesthetic reason is caused by frustration

(either previous, current, or anticipated), an attempt to resist the feelings, a perception of helplessness, and a surrender to feelings of frustration. In both the context of positive and negative emotions it is arguable that tears reflect feelings that cannot be expressed by other behaviours, and may be primarily associated with a feeling of being ‘overwhelmed’

(Vingerhoets & Bylsma, 2015). However, the strength with which these feelings elicit crying is likely influenced by social, cultural, and personal factors, which otherwise can’t be explained by the presence of one or more of these contextual cues.

Universality and Cultural Variation

The lay-understanding that crying is beneficial or helpful in some way is widespread and enduring. This belief is even strongly held cross-culturally (Becht & Vingerhoets, 2002).

Japanese people in particular seem to champion this idea, with a number of organisations providing crying therapy for individuals and organisations under the assumption that, for stress reduction, crying is better than laughing or sleeping, and that crying once a week is the secret to a stress-free life (Jiji Press Ltd., 2018). However, the scientific literature that does exist paints a more complex picture than that which is held by the lay-person.

Although it does appear that crying is often remembered as a positive experience, the THE HISTORY AND THEORIES OF CRYING 23 length of time since crying has occurred influences how people evaluate that experience.

Indeed, over 70% of people report that they generally feel that crying provides positive effects including a feeling of catharsis (Rottenberg, Bylsma, & Vingerhoets, 2008). However, the further away the crying experience is in time, the more positively it appears to be remembered. For example, crying that is reported on the day it occurs, appears to only be endorsed as positive 30% of the time (Bylsma, Croon, Vingerhoets, & Rottenberg, 2011).

Furthermore, experimental results to date where mood is reported immediately after crying episodes, show that crying does not reduce negative feelings. Rather, they appear to worsen mood (Bylsma, Vingerhoets, & Rottenberg, 2008; Gross, Fredrickson, & Levenson, 1994).

The gender and nationality of the crier also appear to influence a person’s frequency of crying and its impact on their mood (Becht & Vingerhoets, 2002; van Hemert et al., 2011).

However, across cultures the same view on the helpfulness of crying seems to be strongly held, with a study finding that of 23 countries, all indicated that on average mood improved following crying (Becht & Vingerhoets, 2002). However, despite this general view of crying as assisting mood change positively, women overwhelmingly responded as crying more frequently than men, with crying occurring three times more among women in some countries. Interestingly, this research found that women’s propensity for crying is exaggerated in many western countries, with women showing considerably greater frequencies of crying compared to non-western countries, where the difference in frequency between men and women is smaller, and in some instances, non-existent. Perhaps unsurprisingly, in more feminine, wealthier countries, where individuals report crying often and with little feelings of shame, people report feeling better following crying compared to countries with opposing characteristics. This suggests a large role played by sociocultural factors that influence beliefs about crying’s utility and actual evaluations of crying (Becht & Vingerhoets, 2002; van

Hemert et al., 2011). Importantly, it suggests a considerable amount of variance in crying CHAPTER 1 24 behaviours and evaluations are influenced by how cultures place value in emotion expression, and crying in particular, dependent on accepted gender roles.

Boys Don’t Cry

Women cry more than men (Vingerhoets, 2013). Before reaching adulthood, however, infants and children of different genders show few (if any) differences in crying frequency

(Vingerhoets & Scheirs, 2000). In fact, this gender difference doesn’t appear to emerge until around the age of 11 (van Tilburg, Unterberg, & Vingerhoets, 2002). Despite speculation that these differences occur because of secondary sex characteristics associated with female development during puberty, namely, menarche (Buchanan, Eccles, & Becker, 1992), these differences seem to be driven by decreased crying in boys during adolescence, rather than an increase in crying among girls (van Tilburg et al., 2002). The role of ‘biology’ has been often presumed to be the cause of this change in crying frequency as we age (see Vingerhoets,

2013), and, though no evidence has yet found support, menstruation has also been largely speculated to be involved in why adult women cry more than men (Horsten, Becht, &

Vingerhoets, 1997; Romans et al., 2017). Very little evidence has been found to support these notions of hormonal changes as responsible for changes in crying across the lifespan, with research instead suggesting that age-related changes in levels of empathy (van Tilburg et al.,

2002) and our social environment (Bronstein, Briones, Brooks, & Cowan, 1996) are most likely to influence these differences.

Across cultures men and women seem to equally endorse the general belief that that they feel better after crying, and yet, contrarily, men associate crying with more feelings of shame (Becht & Vingerhoets, 2002). Many cultures, and particularly Western cultures, value general emotional restraint or stoicism in men (Carlson & Hall, 2011; Jansz, 2000), with emotional competence and intelligence rated more highly when men delay emotional responses. Conversely, these traits of emotional competence and intelligence are seen as THE HISTORY AND THEORIES OF CRYING 25 higher in women when their emotional response occurs immediately (Hess, David, & Hareli,

2016). This is unsurprising given the typical heteronormative perceptions of masculinity as mentally tough, and having emotional reserve, with notions of emotion expression as feminine (Pleck, 1995). This perception of role patterns within a society is perhaps the reason why ratings of shame while crying are higher among men compared to women (Becht &

Vingerhoets, 2002). While we might assume that since crying is tied to ideas of weakness and femininity, men might be judged negatively when crying, that is not the case. In fact, when viewed in emotional contexts where it is appropriate to cry, men have been viewed as more likable, with women, on the other hand, viewed most likable when they do not cry (Labott,

Martin, Eason, & Berkey, 1991). Furthermore, it was found that people who cried were not seen as more feminine. Although there are various and many possible reasons why we find gender differences in crying, it is still unexplained why we engage in this behaviour at all if our responses are so varied, and its benefits unclear.

The Utility of Crying

From ‘purposeless’ to ‘safety valve’, emotional crying in adults has been thought to function in many different ways (Breuer & Freud, 1955; Darwin, 1872). ‘Functional crying’ has been widely debated over the years with the perceived benefits of crying, among both popular media articles and lay-people alike judging overwhelmingly that crying is beneficial for well-being and brings positive outcomes to the crier (Cornelius, 2001; Rottenberg,

Bylsma, & Vingerhoets, 2008). Notwithstanding the paucity of research in this area, many different theories have emerged over the years in an attempt to explain adult crying. The current literature reveals two common overarching functions of crying: (1) the intrapersonal/individual function of crying and (2) the interpersonal/social function of crying. CHAPTER 1 26

Individual Functioning

Catharsis. A number of notions surrounding individual functions of crying have considerable focus on the idea that crying facilitates feelings of catharsis, with many of these theories originating from psychodynamic tradition. Psychodynamic theories in particular have proposed that crying may be a physical symbol for a psychological process of release, and that crying functions as a type of ‘overflow’ representing emotions that have reached a critical level (Breuer & Freud, 1955; Heilbrunn, 1955; Koestler, 1964). A cognitive Two-

Factor Theory proposed by Efran and Spangler (1979) similarly suggests that adult crying only occurs during a ‘recovery’ phase following emotional arousal. Efran and Spangler argued that crying occurs when the cause of crying has disappeared or concluded and signifies recovery rather than continued distress. In testing this theory, some research has found that when watching a tear-inducing film, people reported sadness when a barrier was present in the film, but tearfulness only occurred when that barrier had either disappeared or was resolved (Efran & Spangler, 1979; Labott & Martin, 1988). The Two-Factor Theory, then, indicates that crying is not likely to occur while experiencing distress, but is rather an indication that a conflict has passed.

Unfortunately, both of the psychodynamic and cognitive theories presented fail to explain why we cry when we feel helpless or frustrated since crying is unlikely to occur when distress has been resolved in these situations. Further, as Labott and Martin (1988) argue, both the overflow and two-factor theories fail to predict when other responses might occur

(e.g., laughing). Thus, it seems these theories fail to provide a psychological framework for why and how crying produces this feeling of catharsis, since it is not actually a behaviour, but is the resulting feeling from a release of (emotional) energy. Because of this, others have speculated that crying occurs when other coping mechanisms have been ineffective or are unavailable to the crier (see Vingerhoets, 2013; and Vingerhoets et al., 2000). This has led to THE HISTORY AND THEORIES OF CRYING 27 the more recent hypothesis that crying may serve a soothing function of emotion-focused coping that is likely interconnected with a resulting feeling of catharsis (Gračanin, Bylsma, &

Vingerhoets, 2014; Vingerhoets & Bylsma, 2015). Specifically, self-soothing is a response- focused emotion regulation (emotion-focused) strategy responsible for modulating emotional experience (including physiological arousal that may be associated with emotion).

Self-soothing thus employs a wide-ranging number of potential strategies that can be used to manage mood and specific emotion states. These may be distracting oneself, attempting to relax, or engaging in physical activities. Perhaps in these instances crying may induce self-soothing alone, or may be an intrapersonal signal to oneself to engage in other self-soothing behaviours, such as specific attempts to regulate one’s own emotions via meditation, or externally through music use (Labbé, Schmidt, Babin, & Pharr, 2007; Thoma,

Scholz, Ehlert, & Nater, 2012). The former, in particular, is often demonstrated in people’s reports of why they cry, often stating that people deliberately engage in crying to make themselves feel better (Simons, Bruder, van der Löwe, & Parkinson, 2013). Thus, the literature on self-soothing hypothesises that the self-soothing effects of crying should be part of the consequences of homeostatic regulation encouraged by crying (Gračanin et al., 2014).

Some laboratory research has attempted to test whether crying is related to the arousal or soothing of physiological responses, finding conflicting results across measures of heart rate, respiration, cortisol (stress) responses, skin conductance, and other somatic activity.

These results have found that contrary to the soothing/catharsis view of crying, that crying while watching sad film clips has shown increases in arousal via heart rate, skin conductance, finger temperature, and other somatic activity (for a review see Bylsma, Gračanin , &

Vingerhoets, 2019). On the other hand, respiration rate has consistently shown decreases during crying (e.g., see Gross, et al., 1994), and cortisol also potentially reduces the more a person cries, supporting the soothing/catharsis view (Vingerhoets, & Kirschbaum, 1997). CHAPTER 1 28

Although this research provides some evidence of possible soothing from crying, these conflicting results make it difficult to understand the pattern of soothing. It may be that crying only immediately affects rates of breathing, and perhaps, it takes more time for heart rate and other somatic functions to return to normal but that this may be assisted by crying.

Reductionist perspectives. The idea that crying functions as a form of tension reduction is not new. In fact, Darwin (1872) himself recognised that tears might bring relief to the crier, despite arguing that crying was essentially a meaningless by-product of muscular contractions around the eye. His reductionist view of crying suggests that although, in his view, crying is an exception to the rule of purposeless evolution, that in fact crying may bring relief in similar ways that writhing, teeth grinding, and shrieking, provide relief when a person is in pain (p. 175). Other reductionist perspectives have also similarly argued that crying serves as a protective mechanism for the respiratory system, or to remove biochemical products that are released during feelings of distress (Frey & Langseth, 1985; Montagu,

1959). Likewise, it has also been suggested that rather than toxins leaving the body when crying, crying may instead initiate a release of endogenous opioids within the body, which may have an analgesic and pain reducing effect (Panksepp, 1998).

Problematically, none of these reductionist perspectives have been successfully empirically tested, with only spurious evidence within the scientific literature to date. Indeed, the only observation found in support of these theories was that women’s emotional tears, compared to tears produced by irritants such as onions, contain higher concentrations of proteins, somewhat supporting a notion of an exocrine release of biochemicals through crying

(Frey, Desota-Johnson, Hoffman, & McCall, 1981). There is also some evidence that tears may contain a chemosignal or pheromone that reduces arousal and perceptions of female attractiveness in men (Gelstein et al., 2011). However, neither of the studies were able to show that these changes in the body’s chemistry were actually related to a change in an THE HISTORY AND THEORIES OF CRYING 29 individual’s wellbeing or coping ability. Furthermore, neither of these studies have been successfully replicated, with protein concentrations left un-replicated in the decades since the research was originally published, and chemosignalling effects of tears recently failing to show successful replication (Gračanin, van Assen, Omrčen, Koraj, & Vingerhoets, 2017).

Unfortunately, the idea of biochemicals released through tears during emotional crying ignores the seemingly obvious physical feature of the tear ducts, namely that the majority of our tears do not end up on the face, but drain through the nasolacrimal ducts and actually return to our stomachs (Miller-Keane & O’Toole, 2003). This likely defeats the hypothesised purpose of the ejection of these biochemical distress products. However, none of these assertions rule out the potential intrapersonal signalling function proposed by Panksepp

(1998), where tearing may initiate a change within the body that works to alter internal biological substances to reduce feelings of distress. This raises the possibility that crying may be a self-signalling process that indicates the release of endogenous opioids, which may have a pain-relieving effect on the body. If we assume that much of the personal and difficult emotional experiences encountered that cause crying are ‘emotional pain’ (i.e., a feeling of suffering with an emotional, rather than physical, origin), perhaps Panksepp’s hypothesis explains why we cry when we are in emotional and physical pain.

Social Functioning

Compared to ideas regarding the individual functions of crying, theoretical positions on the social functions of crying are far less well-established, with only very recent attempts to empirically test these explanations. Perspectives on the social function of crying propose that crying is a strong source of communication with others that is maintained from infancy into adulthood (Hill & Martin, 1997). They emphasize that crying is predominantly a visual cue for negative emotions that signal to others that a person is in need of help and succour, as well as providing an unambiguous method of emotion identification (Vingerhoets, 2013). CHAPTER 1 30

Similarly, from the perspective of attachment theory, in infancy and childhood, crying is often a reflexive signal that communicates distress and is primarily designed to influence attention and support from proximal attachment figures (Bell & Ainsworth, 1972).

For adults, similar attentional and communicative patterns occur, with proximal others often responding to crying with support, sympathy, and also sadness (Hendriks, Croon, &

Vingerhoets, 2008; Hendriks & Vingerhoets, 2006; Provine, Krosnowski, & Brocato, 2009).

Evolutionarily, it has been proposed that crying throughout the lifespan mainly serves interpersonal purposes, such as gaining help and facilitating social bonding, which perhaps explains why, at times, we empathically weep with or for others (Gračanin, Bylsma, &

Vingerhoets, 2018). Another evolutionary perspective speculates that crying may inhibit aggression and serve to diffuse violent situations (Hasson, 2009), finding some support in the literature with tears appearing to reduce arousal, testosterone levels, and perceptions of female attractiveness in men (Gelstein et al., 2011). However, these results are somewhat spurious and in need of further research (Gračanin et al., 2017). Others have alternately suggested that crying may be a potent visual cue that adds meaning to otherwise ambiguous facial expressions (Provine et al., 2009). Encouragingly, unlike perspectives on individual functions of crying, empirical research has presented much stronger evidence to support the social function of crying.

The body of research that aims to understand the potential social functions of crying, in particular, tends to support ‘signalling theory’ within an evolutionary approach to crying.

Specifically, signalling theory promotes that specific behaviours evolve because they encourage a reactive behavioural change in an observer, one which would ideally benefit the signaller (Johnstone, 1998). Others have also argued that because signals are generally costly to the individual, they are generally considered reliable or ‘honest’ and that these signals are only displayed if they are in need of favourable responses from the receiver(s), as is generally THE HISTORY AND THEORIES OF CRYING 31 associated with social crying (Kilner & Johnstone, 1997; Vega-Redondo & Hasson, 1993).

On the other hand, children tend to require a threshold of cry-intensity before they are believed to show an honest distress signal (Zeifman, 2001).

Indeed, from experimental and survey findings it is clear that adult crying serves both interpersonal purposes and provides an unambiguous emotional signal. That is, viewing images of crying faces compared to faces without tears have been found to not only make the face appear sadder, but also elicit greater sadness in the observer, more emotional support, less avoidance, and more overall helping behaviours (Balsters et al., 2013; Hendriks et al.,

2008; Hendriks & Vingerhoets, 2006; Provine et al., 2009). Research has found that presenting faces with tears and with tears digitally removed from a face results in increased ratings of perceived sadness on faces with tears. This, of course, may be that the face actually looks sadder, or that the face is less ambiguous in the expression of a particular emotion.

Both of these explanations are in line with signalling theory by increasing observable indicators of emotion states, and also reducing the ambiguity of other possible emotion states.

Perceptions of crying are also affected by other factors that change their intended function. For instance, a recent series of experiments by Vanman, Sharman, and Chantra

(under review) found that male faces appear to show less ambiguity in their emotional display compared to female faces when tears are present on the face. This also appeared to benefit those men, with tears eliciting greater empathy and willingness to help from observers.

Women, however, received the same amount of help whether they cried or not. Similarly, and as mentioned earlier, investigating interactions with a crying person has also found that men seemed to be liked best when they cried to an emotional film, and women liked best when they did not (Labott et al., 1991). Thus, reactions to others crying depends not only on whether they are crying, but also on the gender of the person expressing the emotion. CHAPTER 1 32

As is clear, much of the research on social functions has focused on laboratory investigations of crying, asking others how they view faces with tears or watching others cry, and asking whether they would help. However, this research does not address what actually happens when others are present during an episode of crying and whether this is actually helpful, as the social function of crying would suggest. One survey study found that crying alone or with one other person was related to the most cathartic affects; however, two or more people present when crying were actually related to worse mood (Bylsma, Vingerhoets, &

Rottenberg, 2008). Another attempt to investigate supportive reactions when a confederate cried found that people sympathise more with others who cried, and also cried more in contagion, or possibly empathically compared to when a confederate did not cry (Hill &

Martin, 1997). Although this research is a starting point, they are ultimately not able to address larger questions of whether help from others is actually functional for the crier; that is, whether crying around others results in instrumental help, or if it is merely the presence of others that is comforting. Furthermore, if we are to adopt the social functioning theory of crying, it does not explain why people often cry alone (Bylsma et al., 2008; Vingerhoets, van

Geleuken, van Tilburg, & van Heck, 1997).

A Multifunctional Perspective

Although these earlier functional theories on crying attempt to account for either the biological, psychological, or social processes that may be involved in crying, they ignore much of the overall individual response. Specifically, a person’s crying response involves an interplay of social, cognitive, behavioural, and physiological experience. Most perspectives on crying have tended to narrowly focus on one or two of these aforementioned aspects but have failed to provide a comprehensive theory incorporating each of these processes.

However, newer theories have attempted to combine the broader individual and social theories into more comprehensive models, which consider that crying occurs when a person THE HISTORY AND THEORIES OF CRYING 33 is unable to cope with a situation and can use both inter- and intra-personal resources to aid coping (Miceli & Castelfranchi, 2003). That is, crying can serve as an independent way of self-soothing and regulating distress when feeling helpless, or as a way to gain resources from others who may work to help solve the problem the crier feels powerless to resolve themselves, or just to provide comfort and support (social-soothing). Indeed, there is a differentiation made among criers of the value of crying independently compared to crying when surrounded by others. However, it is clear that there is very little research available able to comprehensively answer these questions. Although some models have been proposed regarding socially mediating effects of crying on mood (Gračanin et al., 2014), only one has attempted to consolidate both positions on crying (individual and social; Vingerhoets et al.,

2000). However, this model, once again, does not focus on individual factors, such as self- soothing that appear to be involved in how crying may affect emotional states.

To attempt to reconcile these themes within the literature, I have developed a parsimonious multi-functional model of adult crying (see Figure 1). This working model incorporates both the individual and social components of crying and how they may function individually or in tandem to affect mood, while also incorporating some of the possible influencing factors of crying that have been mentioned in this review. Based on the previous research outlined above this model proposes that crying leads to social-environmental and/or individual person-centred responses that may co-occur depending on the social context. A social response is defined as a person (or possibly animal) who is present during crying . This person may provide help to a crier, which could be instrumental or emotional, for example, providing tissues, a hug, or providing a solution to a problem. Individual focussed responses may involve self-soothing behaviours and external emotion regulation behaviours such as drinking tea, meditating, or listening to music. These pathways lead to a criers evaluation or re-evaluation of their emotional state and the causes of crying, which will likely lead to some CHAPTER 1 34

Figure 1.1. A new multifunctional model of crying

emotional change. Of course this evaluation may result in feeling better or worse.

These responses to crying and the resulting evaluation are likely to be influenced by different factors related to the crier and the social context. These may be biological factors, with crying potentially impacting stress levels, self-soothing, and a hypothesised return to homeostasis; cultural and social/gender norms that may impact how crying is perceived

(helpful vs. weak); and/or individual differences in how crying is viewed or personality differences that may impact how crying is perceived and evaluated. These cultural and social norms would likely also influence how others respond to the crying of others.

Testing the Multi-functional Model: An Overview of the Thesis

The aim of this thesis was to more comprehensively explore our understanding of adult crying and how it functions in individual and social settings to help (or hinder) recovery from negative emotion states, through testing elements of the multifunctional model of crying.

Although there have been many speculative theoretical positions on crying, investigations THE HISTORY AND THEORIES OF CRYING 35 into this phenomenon have not always been consistent or methodologically rigorous. Thus,

Chapter 2 of this thesis outlines methodological considerations in crying research in outlining how to empirically investigate crying, and its limitations within the field.

Chapter 3 addresses the hypothesised biological factors involved in crying, which are a possible moderating factor in the multi-functional model presented in Figure 1. Thus, the aim of this chapter was to understand if crying involves a protective self-soothing effect via a return to or maintenance of homeostasis following stress as part of the individual responses pathway in the model. The use of psychophysiological methods in this chapter also provides some insight into how biological factors are involved in ‘individual reactions’ during crying.

Overall, this research is a comprehensive psychophysiological investigation into theoretical perspectives on crying as aiding physical and emotional recovery, as well a first effort to understand if crying acts as a barrier to future pain.

The next chapter reports on the development and psychometric validation of the

Beliefs about Crying Scale, a new measure assessing beliefs about whether crying leads to positive or negative emotional outcomes in individual and interpersonal contexts. This research aimed to test a proposed moderator involved in mood change following crying, and to further understand the different possible beliefs and appraisals people generally hold that may influence mood change. This chapter focuses on the appraisal section of the model in understanding what individual difference factors may affect how people evaluate and understand their crying experiences.

Next, Chapter 5 investigated the role of gender, culture, and social reactions to crying.

This was a large multi-country study that aimed to (1) investigate how beliefs about crying are involved in people’s actual evaluations of crying and whether beliefs about crying are more predictive than their self-ascribed gender attributions of crying behaviours and mood following crying. Furthermore, this research also aimed to (2) understand how the presence CHAPTER 1 36 of others affects our evaluations of emotion following crying. This chapter, thus, also investigates whether crying around others really does lead to emotional and/or instrumental help from others. Overall, Chapter 5 focused on again testing the appraisal section of the model, and also the social responses pathway to gain a greater understanding of the whether help is received from other people in social contexts and if that help (or lack of) relates to a change in emotional state.

The final chapter concludes by re-exploring The Multifunctional Model of Crying based on the research contained in this thesis. Here I discuss whether this research provides support to the model and where further research is still needed. MEASURING AND INDUCING CRYING 37

CHAPTER 2: Inducing and Measuring Crying: A Methodological Review

CHAPTER 2 38

Crying is a multi-faceted form of emotion expression that can and does occur in a wide range of contexts and for a wide range of reasons. One can cry alone or in public, out of pride, laughter, anger, sadness, embarrassment, or from physical pain. Crying has also played a significant part of our interpretation and responses to aesthetic pieces, such as music, art, and opera, with these experiences often identified as ‘aesthetic crying’ or a feeling of being moved to tears (Cotter, Prince, Christensen, & Silvia, 2018). Because of the considerable number of antecedents and, as was noted in Chapter 1, the likely large influence our current and formative socio-cultural environment has on our expression of crying (Bronstein et al.,

1996), attempts to scientifically explore the phenomenon of crying in adults have been challenging.

Many different methodological approaches have been used to understand crying in social and personal situations, its relationship with physiology, and to generally understand why crying is considered to be helpful. Unfortunately, definitions that help to guide research have been mixed in the literature with varying emphasis on vocal crying, where vocal crying can occur in the absence of tears (Green, Whitney, & Potegal, 2011; Provine, 2012); crying from distress versus other emotional cause, such as happiness; and even crying as determined by neurophysiology and culture (Bylsma, Gračanin, & Vingerhoets, 2018; Vingerhoets, 2013).

Given the range of definitions on crying - how we induce and measure crying have been particularly difficult hurdles among attempts to understand it. Consequently, this chapter provides a review of methodological approaches within the crying literature to offer a broad overview of previous attempts and techniques to understand crying.

Who Cries and Why

Although crying is a human universal, a particularly important and large part of the crying literature focuses on the demographics of criers. In particular, research on crying has MEASURING AND INDUCING CRYING 39 been focused within social and cultural differences between criers and the gendered differences in crying frequency between men and women. This research has most often utilised quantitative and qualitative methods to gain a more nuanced understanding of who cries and its causes. From this research, we know that across cultures and overall, women cry more often and are more prone to crying than men (De Fruyt, 1997; Denckla, Fiori, &

Vingerhoets, 2014; Fischer et al., 2004; van Hemert et al., 2011; Vingerhoets & Scheirs,

2000). This consistent finding has affected the way in which we approach crying research, with investigations into crying largely focused on the crying experiences of women and attempts to understand why women cry more, with fewer attempts to explore why men cry less (but see MacArthur & Shields, 2015). This is particularly interesting given the finding that gender differences emerge around the age of 11 with males crying less as adolescents, and women’s crying frequency remaining relatively stable (van Tilburg et al., 2002).

It has been speculated that changes to crying frequency around adolescence may be due to physiological changes (hormonal changes and/or the onset of menarche in girls), psychological changes (empathy and personality), and/or influences of the social environment that occur around this age (van Tilburg et al., 2002; Vingerhoets et al., 2000).

Notably, despite historically long and frequent speculation on ‘why women cry more’, no link has been found between menstruation and crying frequency (Horsten et al., 1997; Romans et al., 2017; van Tilburg et al., 2002). Curiously, other possible hormonal links have yet to be explored, such as changes to production in testosterone in men, as a potential explanation to

‘why men cry less’, beginning around adolescence. On the other hand, social factors, including adherence to gender roles, appear to play a large part in how much we cry (Becht &

Vingerhoets, 2002; Bronstein et al., 1996). In fact, one study focussed on men’s crying behaviours found that sex-role adherence to traditional masculinity resulted in less expressions of sadness through crying, compared to men who adhered less to these traditional CHAPTER 2 40 sex roles (Ross & Mirowsky, 1984). Of course, other factors may also impact the expression of crying, with the interpretation of social judgement likely playing a role in whether crying is expressed in certain situations. Specifically, perceived judgement and corresponding shame appear to be felt more intensely by males after crying in social situations (Bylsma et al.,

2008; Simons et al., 2013).

Nonetheless, much research has remained focused on women’s crying, in part because women are often easier to collect data from via inducing crying or examining more frequent crying behaviours, and because they are less likely to be influenced by perceived judgement and shame about crying in social contexts. This has predominantly influenced the way in which we approach laboratory research on crying and its effect on the crier. Furthermore, this continues to affect how we discuss crying, as primarily significant for women as overly expressive of emotions, and not necessarily recognising men as overly restrictive in their emotional displays.

Research Methods

Research conducted on crying has employed both qualitative and quantitative methods to better understand why we cry and what happens when we cry. Qualitative methods, in particular, have been instrumental in gaining a richer understanding of the reasons we cry, (Miceli & Castelfranchi, 2003) and how we feel about crying (Sharman,

Dingle, & Vanman, 2018) through interview, focus groups, or qualitative survey design.

Quantitative research methods have also been utilised extensively throughout the literature to better understand the consequences associated with crying at individual and social levels.

Ideally, the goal of quantitative research is to experimentally identify systematic changes to an observable phenomenon by empirically measuring changes surrounding that phenomenon

(Henn, Weinstein, & Foard, 2006). To observe changes and be able to attribute those changes to a causal factor is one of the primary reasons quantitative methods are used in research. MEASURING AND INDUCING CRYING 41

Consequently, experimental quantitative research where randomisation can be effectively used has been highly informative in attempts to understand the impact of crying thus far and has been particularly successful in research on the social function of crying. However, given the nature of crying, when trying to understand the impact of physical tears on the person who is crying, participants cannot be randomly assigned to crying or non-crying conditions.

Problematically, this means research using experimental methods where crying is induced have, thus far, always been quasi-experimental designs, reducing our ability to make causal inferences about the direct effects that crying may have.

Crying research using a mixture of qualitative and quantitative measures has also been commonly employed (Becht & Vingerhoets, 2002; Sharman et al., 2018; van Hemert et al., 2011). Generally, these mixed methods have been utilised in survey designs that rely on retrospective self-reports. Although surveys have been instrumental in contributing to our understanding of crying, they are also vulnerable to bias from participants through self- presentation and self-protection bias, and memory distortions. This is problematic for crying memories in particular, where people may report stereotyped memories of a crying event, or perhaps retrieve a memory distorted by emotional intensity (Levine & Edelstein, 2009).

Notably, from the existing research on crying it is clear that memory of a crying event is influenced by the passage of time. Assessments made in close proximity to the crying event tend to yield reports of worse emotion states compared to before crying. Meanwhile, assessments made more distantly in time tend to result in reports that crying led to feeling better following crying, suggesting crying might lead to a positive emotional change (Bylsma et al., 2011, 2008; Gračanin et al., 2015; Rottenberg, Bylsma, & Vingerhoets, 2008). Thus, research in this area is often conflicted, with reports of mood or emotional state following crying potentially influenced by memory distortions or bias. This means that research CHAPTER 2 42 measuring crying should be sensitive to time and how the passage of time may affect how people view their crying experiences.

Measuring Crying

Retrospective Self-report

In order to effectively explore adult crying, researchers have made attempts to measure crying behaviours; however, it has been often difficult to know how robust some of these measures are. Most commonly, crying has been measured through self-report simply by asking whether a person cries or not in certain situations, or in asking how much a person cried in their last crying episode (Williams, 1982). In exploring the crying experience this has also meant researchers often just query a person’s frequency of crying over the last month and their mood following crying experiences, which has some benefits, but a number of pitfalls if memory distortions make these episodes difficult to remember in detail (e.g., Becht

& Vingerhoets, 2002). Others have more readily attempted to create self-report scales that measure the crying experience. This has resulted in only two1 considerable published scales in the literature thus far, one being the Beliefs About Crying Scale (BACS; Sharman et al.,

2018) that measures evaluations of crying’s helpfulness when alone or in public and is reported in Chapter 4 of this thesis. The other is the Crying Proneness Scale (CPS; Denckla et al., 2014), which measures how susceptible one is to crying because of different antecedents.

The CPS is an important facet of measuring crying that goes over and above that of simply measuring the actual frequency of crying (Denckla et al., 2014). It is important to note that crying frequency, although a seemingly valid measure of crying, may not provide a true overview of a person’s propensity to cry if they have not encountered cry-eliciting situations over a particular period. For example, a person may be likely to cry when experiencing loss,

1 There have been other crying scales prior to these; however, they appear to have been created out of necessity and not via theoretical and systematic guided development so are not reported here (see: De Fruyt, 1997; Labott & Martin, 1987; Vingerhoets & Becht, 1996) MEASURING AND INDUCING CRYING 43 interpersonal conflict, or when feeling touched at a happy occasion such as a wedding; however, they may not experience any of these events over short periods of time, so asking about crying frequency over the last month may result in lower reports of crying than a person’s actual proneness to cry at any point. Therefore, the CPS provides a valuable measure of overall crying propensity that exists regardless of whether cry-eliciting situations are encountered. However, this measure only accounts for proneness to cry in contexts that are external to the person reporting, such as ‘seeing someone hurt’ and not to events that a person is directly involved in, such as ‘fighting with your partner’.The items on this scale may or may not capture events that are personally relevant to the person. For example, the person who is hurt could be known to the crier or they may be someone they do not know. However, the items only refer to external events that have no information regarding the direct impact of an event on a person. Although the CPS is generally good at estimating proneness to cry, it is important to recognise that proneness may be different when events are personally affecting.

Naturalistic Self-report. Other intrapersonal investigations into crying have employed longitudinal research designs to understand how crying responses differ over time and if there are corresponding physiological changes that may be related. Unlike most questionnaires on crying that tend to collect more global data at a single point in time (or used before and after inducing crying), diary studies are able to record multiple points of data across time. This design allows researchers to better understand how different variables might change across time, such as mood, and see how they might be affected by whether a person cried or not. Diary studies have been instrumental in gaining an understanding of women’s crying behaviours across time and in naturalistic settings with a primary focus on crying’s relationship with menstruation (Romans et al., 2017; van Tilburg, Becht, & Vingerhoets,

2003). These studies asked participants to track both their menstrual cycle and their crying behaviours, not only asking their frequency and proneness to cry each day, but also how CHAPTER 2 44 intense their crying was, how their mood changed following crying, and the time of day they cried (Bylsma et al., 2011). This research has been influential in revealing the weak relationship between women’s crying and menstruation, and by using naturalistic data collection methods, have also informed a richer understanding how crying functions in every- day life.

In the Laboratory

In exploring the effect of crying on the crier and those around them, several experimental and non-experimental studies have been conducted throughout the last 20 years.

These studies have attempted to understand both how others react to crying and what happens to an individual when they cry. Studies focused at an intrapersonal level have been particularly determined to understand what happens within the body when we cry and has led to a number of studies exploring physiology that may be related to crying. This research focus has been encouraged by hypotheses that crying works to calm and soothe the body and the mind, either by providing a physical or psychological release that may be facilitated by biochemicals (Frey & Langseth, 1985; Panksepp, 1998).

Physiological laboratory studies that have investigated this concept have studied a range of physiological mechanisms that may be related (see Chapter 1, ‘Catharsis’, for general findings on physiology). Most of these have centred around the measurement of respiration and heart rate, with variations in the measurement of skin conductance, blood pressure, temperature, and hormones such as secretory immunoglobulin A or cortisol (for examples, see Gross et al., 1994; R. B. Martin, Guthrie, & Pitts, 1993). More recently, research has begun to focus around the measurement of facial thermography during crying

(Ioannou et al., 2016). That is, measuring temperature changes across the face, with researchers investigating whether crying displays a distinct thermal (blood flow) pattern in the face. MEASURING AND INDUCING CRYING 45

Laboratory research such as those mentioned above generally involve a baseline period to gauge general autonomic functioning followed by a negative mood induction to encourage crying in participants where their physiological responses are measured and then assessed for changes in functioning compared to baseline and/or a pre-crying period. In my own research (see chapter 3 of this thesis) this same paradigm was used to measure physiological changes when crying. Participants were asked to sit for a baseline period to measure heart rate, respiration, and cortisol. They were then exposed to a stimulus to induce crying while heart rate and respiration were recorded and another cortisol sample was taken following crying. To further test the functional theory that crying may be analgesic and assist in coping with stress and pain, we measured the effects of crying by asking participants to hold their hand in an ice bath for as long as they could while also measuring continuous changes to heart rate, respiration, and measure cortisol changes again.

Social Context

Crying is primarily theorized as a communicative signal, and as such research on how we respond to others’ crying has been abundant, with experimental and survey investigations into this phenomenon. Experimental research in particular has provided valuable insight not only showing that when observing faces with tears versus (the same face) without tears, people rate crying faces as sadder, whether the tears are presented on a face that looks happy or sad (Fitzgerald & Vanman, 2018; Provine et al., 2009). Similar research using faces such as these have tested the speed at which we recognise tears on faces by rapidly presenting faces to participants for 50 milliseconds (ms) and then asking them to rate the extent to which they felt the person in the image required social support (Balsters et al., 2013). Results from this research showed that faces with tears on both sad and neutral faces were rated as needing more social support than faces without tears. Overall, this research presents some evidence that we may unconsciously process tears, given the speed at which these images were CHAPTER 2 46 presented. However, there was no masking stimulus on the image trials to reduce conscious processing of the faces after presentation. Without a mask, visible persistence effects can last another 100-200 ms following the onset of the stimulus (Coltheart, 1980), so it is unclear if unconscious image processing of tears actually occurs this quickly. Nevertheless, these ratings of social support have been found in several other studies where images were presented for longer periods (Hendriks & Vingerhoets, 2006; Provine et al., 2009; Zeifman &

Brown, 2011).

Research investigating differences in responding to images of faces with or without tears have also utilised physiological methods to infer others’ responses to perceiving tears.

Some have utilised facial electromyography, measuring minute muscle movements of the face linked with emotions and facial mimicry, finding that people unconsciously frown when they see tears, and, surprisingly, that this frowning response occurs whether the emotion on the face is sad or happy (Fitzgerald & Vanman, 2018; Sharman, Vanman, & Scambler, 2015).

Another innovative investigation using psychophysiology to measure social responses to crying collected women’s (sad) emotional tears or a salt solution and applied them under male participants’ noses to investigate changes in testosterone and arousal (Gelstein et al.,

2011). Following the application, smelling women’s tears resulted in testosterone reduction, reduced sexual attraction to women’s faces, and reduced sexual arousal compared to those who smelled saline solution only. Although these results are quite remarkable, to date, they have yet to be successfully replicated. Indeed, one attempt to replicate these findings showed no evidence that women’s tears reduced male testosterone, sexual arousal, or sexual attraction to women’s faces (Gračanin et al., 2017). This extended replication also found no evidence that emotional tears were better than tears produced by irritants (e.g., onions) in increasing pro-sociality or lowering aggression levels. While these results are intriguing, the MEASURING AND INDUCING CRYING 47 methodology used to gather this information is highly novel and provides new and innovative ways to test the effects of crying and its effects on the body.

Less common in research on responses to crying are naturalistic explorations that expose people to crying others. Only two studies have used such a paradigm. Both of these studies involved movie watching scenarios where participants and confederates were seated together with the confederate crying or not crying during the movie (Hill & Martin, 1997;

Labott et al., 1991). Participants in both studies were asked to rate their attitude and reactions to the crier. However, in one, empathic weeping was measured through self-report.

Specifically, researchers investigated how much participants cried when a confederate was crying, compared to their crying intensity when the confederate did not cry (Hill & Martin,

1997).

Outside of more complicated laboratory research on the social context of crying, some survey research has also been conducted. These self-report, often single time-point studies

(but see Bylsma et al., 2011) have focused more on the person who cries and how they feel dependent upon how many people are present. However, this research has unfortunately been quite limited and thus far has only provided a basic understanding of mood change potentially worsening when more people are present during crying, but not potential help received from those persons (Bylsma et al., 2008; Bylsma et al., 2011).

Of course, these investigations are still not able to identify what happens in a real- world context, where crying often occurs because of nuanced and personal difficulties, such as loss or interpersonal conflict, and whether these attitudes and reactions would be similar in those circumstances. Specifically, these findings provide no insight into whether if around another person who is crying someone actually provides help and if that help is perceived by the crier to be helpful, particularly with regard to mood improvement. Furthermore, most of the studies reported here most often consider crying as part of an emotional state of sadness, CHAPTER 2 48 with laboratory explorations of crying often centred around the idea of ‘sad-crying’, or specifically crying because of emotionally sad events or stimuli. This focus on sad-crying makes generalisability of these results difficult when considering the number of different emotional reasons for crying.

Inducing Crying

Given that many intrapersonal investigations into crying rely on a participant to cry in order to measure physiological and psychological changes following crying, researchers often induce crying in participants. Various approaches have been utilised across the literature to induce crying, with some researchers asking participants to remember sad personal events

(Delp & Sackeim, 1987), using empathic crying, as described above (Hill & Martin, 1997), and non-emotional tears have also been induced via irritants, such as onions (Frey &

Langseth, 1985; Gračanin et al., 2017). However, in laboratory research emotional crying is most often induced through watching videos or films that are based around emotionally sad stimuli. These videos are often long (up to 70 mins; Hendriks, Rottenberg, & Vingerhoets,

2007), and generally centred around tragedy or trauma (see Chapter 3, Table 3.1 for emotion induction techniques and timings used in publications).

Rarely are shorter emotion induction videos used in inducing crying, with few published attempts at using shorter videos in the literature. Of the two that have attempted to use shorter cry-eliciting videos in research (3:25; 2:50 mins), they have, at best, had only

28% of participants respond with tears, suggesting that perhaps shorter videos are not as effective in cry-induction as longer videos that require participants to engage for much longer in a story (Gross et al., 1994; Rottenberg, Gross, Wilhelm, Najmi, & Gotlib, 2002). Watching others experiencing distress or social pain interacts with our own direct experience of pain, which may explain why we cry to the suffering of others (Jackson, Meltzoff, & Decety, 2005;

Mischkowski, Crocker, & Way, 2016). Thus, perhaps shorter videos have been less effective MEASURING AND INDUCING CRYING 49 in inducing crying because there is not enough time to connect and empathise with those in the video, compared with longer video exposure. However, given the extent of short videos used to capture attention online, I investigated the effectiveness of short videos used as crying induction methods as part of preliminary research into crying compared to previous research that has used short excerpts of films.

Fifty-two participants completed an online study in which they randomly viewed six different (happy or sad) emotional video stimuli. The stimuli included a short excerpt from the film ‘The Champ’ (video 1), widely used to elicit ‘sad’ mood, and previously used as a short cry-induction method (Rottenberg et al., 2002); the film sequence from the beginning of the animated film ‘Up’ (video 2); a ‘happy’ video clip featuring a returning soldier surprising her son (video 3); a happy news report featuring a heart transplant recipient meeting organ donors family (video 4); a sad clip about a dog’s last day before it is put down (video 5); and a both sad and happy advertisement involving a father and daughter’s relationship (video 6).

After each video participants were asked if they had seen the video previously and asked following each video what their “most intense emotional experience” was while watching

(1= no reaction, 5= choked up, 9= sobbing). This item measures self-ratings of crying and has been used previously in measuring crying responses to films (Labott, Ahleman, Wolever,

& Martin, 1990).

Crying frequency to each video was recorded as ‘crying’ when participants reported any type of tearing, corresponding to ratings 7 and above on the 9-point scale. Specifically, when participants recorded their emotional reaction as tears in eyes, tears down face, or sobbing. All videos induced crying responses in at least some participants; however, some were more effective in inducing crying in a greater number of participants. Interestingly, the short film excerpt from The Champ had the lowest number of tearful responses (see Table

2.1). However, (video 5), which was a stand-alone short video documenting a dog’s final day CHAPTER 2 50 before being put to sleep had the most tearful responses with 46.2% of all participants responding with tears and the number of cry responses much higher for women than men.

These results may indicate that short videos, rather than excerpts from films are more effective at eliciting crying. The use of excerpts might not be effective for eliciting such strong emotional responses if the original film or video required more time to empathise with the characters, which may be the case for The Champ, whereas shorter videos attempt to encompass the entire story in less time and don’t require more context or time to develop character stories. Furthermore, these results suggest that sad videos are more effective than happy videos in inducing emotional crying in participants. This may be because sad emotional stimuli are more provoking and closer to the events we cry to in our personal lives.

The ability to induce a high frequency of crying responses through short videos is an important development for future research. Particularly given previous research has not been able to adequately induce crying in a large number of participants, with the largest sample size of 150 only able to induce crying in 33 participants using a short film excerpt (Gross et al., 1994). This is problematic with many previous studies that often already use small sample

Table 2.1: Proportion of criers to short videos and film excerpts Total Criers Males Females The Champ (video 1) 11.5% 12% 11% Up (video 2) 23.1% 18% 32% Soldier surprise (video 3) 15.4% 18% 32% Heart transplant (video 4) 23.1% 12% 42% Dog’s last day (video 5) 46.2% 37% 74% Father/Daughter (video 6) 28.8% 24% 37%

NB: Women reported significantly more crying responses to the soldier (t(23.44) = 2.17, p = .04), dog (t(50) = 3.26, p = .002.), and heart transplant video (t(27.02) = 2.31, p = .029), with no differences between groups for any of the other videos. There was no relationship between previous exposure to the video, except for video 4, with those who had seen the video before more likely to report emotional responses to the video (r(52) = .32, p = .02). MEASURING AND INDUCING CRYING 51 sizes, meaning that research conducted in this area, especially in laboratory research, uses small and/or uneven samples of criers compared to non-criers (see Table 3.1). Enhancing the methodology in this area will improve the inferences made from crying research.

Nevertheless, although videos are useful to induce crying in laboratory settings, they may not elicit the same kind of crying that occurs naturally, and which are generally prompted by personal life-events. Regardless, this is the most effective and efficient cry-induction technique available, particularly for laboratory studies.

Conclusion

What is clear from this methodological review is that the approach to research on crying has most often been investigated from the perspective of women. Although for many laboratory investigations in this area this has been a practical sampling strategy, it has also meant that we have failed to investigate the impact of gender roles and how they may influence our behavioural expressions of crying from sadness. Also ignored within this literature is a focus on social help that a crying person receives. Indeed, although research has focussed on how others report they might respond to someone else who is crying when looking at images of crying faces, we can be less sure about whether this effect translates into behavioural responses in the real world. This means that while we know that people report that they are more likely to provide help and resources to someone who cries, there is a dearth of literature informing us about what a crying person’s experience is when others are present during crying.

This review also highlights the problems in the measurement of crying and the different facets that might impact how we engage in crying. Specifically, the one previous formalised crying measure. Although this measure is important in providing a measure for proneness to crying, it focuses solely on causes that are external to the crier. This leaves us without a formal way of measuring how personal or internal causes of crying impact the crier, CHAPTER 2 52 not whether these internal causes follow the same crying proneness patterns as external causes.

Finally, although the nature of this research means that laboratory research is approached quasi-experimentally, studies have faced problems in appropriate induction of crying in participants, and appropriate comparison groups within the research because of this.

That is, laboratory research faces issues with interpretation without highly powered and well- designed methods to induce and measure crying across those who are exposed to sad stimuli, and those who are not. In doing so, researchers can better inform an understanding of how crying influences our physiological and emotional responses. RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 53

CHAPTER 3: Using Crying to Cope: Physiological Responses to Stress Following Tears of

Sadness

This chapter is from a manuscript in press at Emotion

Contribution to authorship:

The conception and design of the project was completed by myself (75%) and Dr. Vanman

(20%). Dr. Vingerhoets provided feedback on the early design of the project (5%). All data was collected and analysed by me, with some data collection assistance from two research assistants (5%), Renée Calligeros, and Hannibal Thai. All important theoretical, analysis, and journal revision decisions were made by me. The entire manuscript was drafted and edited by me with review provided by all co-authors. CHAPTER 3 54

Abstract

This research tested the hypothesis that emotional crying facilitates coping and recovery, specifically through physiological changes that occur during crying. Female undergraduate students (N=197) were randomly assigned to either a sad or neutral condition using short videos. Sad videos were selected for their extreme emotion elicitation. We predicted that compared to those who did not cry to the stimuli and those who were exposed to neutral videos, people who cried would (a) be able to withstand a stressful task for longer;

(b) show lower levels of cortisol following crying and exposure to the stressor; and (c) have faster recovery (i.e., return to baseline levels of affect). The final groups consisted of the neutral group (N=65), sad criers (N=71) and sad non-criers (N=61). After a 5-minute baseline period, participants watched either the sad or neutral videos for 17 minutes and then completed a physical stressor (cold pressor test). Heart rate and respiration were continuously recorded while salivary samples for cortisol were taken at four separate time points during testing. Analyses revealed no differences between the three groups in time withstanding the stressor or cortisol changes. Respiration rate, however, increased in the neutral group and non-criers while watching the videos, with criers’ respiration remaining stable. Furthermore, heart rate was found to decelerate just before crying, with a return to baseline during the first crying period. These results suggest that crying may assist in generally maintaining biological homeostasis, perhaps consciously through self-soothing via purposeful breathing and unconsciously through regulation of heart rate. RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 55

The question “why do we cry?” has captured human curiosity throughout history, from religion and literature to scientists alike. Despite concluding that tears might bring relief to the crier, Darwin (1872) posited that emotional tearing served no purpose, arguing that crying was a meaningless byproduct of muscular contractions that work to protect the eye.

However, crying can take different forms and be accompanied by changes to facial expressions and vocalisations, as well as sobbing (convulsive inhaling and exhaling of air).

Physically, of course, crying is simply characterised by tears shed from the lacrimal apparatus, often due to irritation. Yet emotional crying can occur without the irritation of ocular structures (Patel, 1993), with the relationship between crying and mood or emotions still unclear. Indeed, the functions of emotional crying, which we refer to here as any shedding of tears from the eyes as a response to sadness, has been widely debated in both the scientific literature and media.

There is an evident saturation among popular literature on the effects of crying as a positive expression and release of emotion, with one study finding that 94% of magazine articles on crying over 140 years recommended crying as beneficial for one’s well-being or as

‘healthy’ (Cornelius, 2001). In fact, many of these articles even claim that the suppression of one’s tears negatively affects wellbeing. Current scientific literature paints a more complicated picture with some studies finding that crying episodes caused by negative affect sometimes do, and sometimes do not, induce self-soothing (for a review, see Rottenberg,

Bylsma, & Vingerhoets, 2008). Despite some competing theories and results within research on crying, existing literature on the less recognised interpersonal (social) functions of crying, to date, are relatively robust. For example, research on these social functions have found that when viewing a sad crying face, people are more likely to experience feelings of sadness, less likely to report feelings of avoidance, and more likely to provide emotional support and helping behaviours overall (Balsters et al., 2013; Hendriks et al., 2008; Hendriks & CHAPTER 3 56

Vingerhoets, 2006; Vingerhoets, 2013; Vingerhoets, van de Ven, & van der Velden, 2016).

However, research concerning the intrapersonal (individual) functions of crying has yielded less consistent findings, with effects varying depending on the timing of questioning, context of crying, and affective state of the crier (Bylsma et al., 2011, 2008; Rottenberg, Bylsma, &

Vingerhoets, 2008; Rottenberg, Bylsma, Wolvin, & Vingerhoets, 2008).

How does crying make us feel?

Does crying make us feel better when we are sad and do people who cry feel better than sad people who don’t cry? Differences in the methodology of retrospective studies have made it difficult to interpret the literature. Variables such as the characteristics of the crier, recency of the crying event, and what caused the crying, appear to affect the evaluation of its effects on mood and how observers react (Rottenberg, Bylsma, & Vingerhoets, 2008).

Despite this variation, it is clear that generally people remember crying episodes as helpful for mood or emotional improvement. In particular, over 70% of people report that crying brings some form of psychological benefit, often reported as a feeling of catharsis when they are asked generally how they felt after previous crying episodes (see Vingerhoets, 2013).

However, when asked to think specifically about their last crying episode, this reduced to

51% of people reporting that crying brought positive effects (Bylsma et al., 2008). Further, when women were asked on the day the crying occurred, only 30% reported positive effects of crying (Bylsma et al., 2011).

When crying has been induced in laboratory settings, rarely have people reported feeling better immediately after crying. In fact, criers always report feeling much worse than before, and sometimes worse than those who watched the same stimuli, but did not cry (see

Cornelius, 1997 for an overview; Gross et al., 1994; Rottenberg, Bylsma, & Vingerhoets,

2008). However, other studies have reported no differences between these subgroups, RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 57 complicating the assumption of emotional intensity that crying is assumed to have over and above that of sadness alone (Hendriks et al., 2007; Kraemer & Hastrup, 1988).

Aside from increased subjective negative mood following crying, criers have also generally shown increased somatic and autonomic arousal with changes in heart rate (HR), skin conductance, finger temperature, and other somatic activity (for further review, see:

Bylsma et al., 2018; Gross et al., 1994; Hendriks et al., 2007; Kraemer & Hastrup, 1988;

Labott et al., 1990; Rottenberg et al., 2002). These results support the notion that crying actually produces a state of high physiological arousal, rather than the idea of crying as a recovery process. Respiration rate has, however, shown consistency with the popular recovery hypothesis with most studies finding that those who cry show decreases in respiration rates compared to those who do not (Gross et al., 1994; Hendriks et al., 2007).

That is, breathing slows down when crying, rather than speeds up, which tends to occur when aroused. In particular, one study found that although HR increased during the onset of crying, this increase was only maintained for the one-minute period after crying began. Conversely, respiration rate showed an extended period of reduction continuing for at least 3 minutes post-crying onset (Hendriks et al., 2007). Other preliminary data have also suggested that the amount or intensity of the crying may be associated with reductions in cortisol levels, but the cortisol collection in this study accounted only for diurnal patterns, not for food and liquid intake, medication, and exercise (Vingerhoets & Kirschbaum, 1997).

These studies provide some evidence that crying may indeed be linked to recovery following distress. Perhaps we see these differences because effects from crying may not immediately occur, but may take time. That is, immediate measures of affective change may not reflect the true impact of crying on individual recovery from stressors. To date, only one study has investigated mood changes over a longer period following crying, finding that, indeed, mood improves after a short amount of time for both criers and non-criers, but CHAPTER 3 58 appears to improve even further only for those who cried at follow up 90 minutes later

(Gračanin et al., 2015). Notably, however, by definition all studies suffer from the limitations of a quasi-experimental design with in most cases a limited number of participants who actually cry and with no neutral condition for comparison – thus only comparing people who cry and who don’t cry in response to a sadness induction in these designs. This lack of a neutral comparison group means that previous research has been unable to conclude whether there may be other systematic differences between criers and non-criers to sad stimuli.

The overall interpretation of findings in the crying literature, especially that of laboratory investigations that elicit crying, is still equivocal. Despite some laboratory studies recruiting over 100 participants comparing criers to non-criers, these larger studies have only managed to elicit crying in 22% of participants, while also neglecting to include a neutral comparison group (see Table 3.1; Gross et al., 1994; Rottenberg et al., 2002). In fact, of the laboratory studies that have reported the proportion of participants who cried during exposure to a sad stimulus, the total number of criers range from as low as 6 or 8 to as high as 33, with the number of non-criers in comparison often much higher, with one such study reporting up to 117 non-criers (Gračanin et al., 2015; Gross et al., 1994; Hendriks et al., 2007; Kraemer &

Hastrup, 1988; Labott et al., 1990; Marston, Hart, Hileman, & Faunce, 1984). These often small and unequal sample sizes undermine the reliability of the current crying research using emotion elicitation procedures to induce crying, with the possibility that results of some of these investigations may not reflect true effects with potential overestimates of effect size and low reproducibility. Thus, it is important to continue to investigate the physiological and psychological changes that occur among criers to strengthen the reliability and validity of research within this field. RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 59

Table 3.1: Summary of cry-induction laboratory studies and overall numbers of criers analysed in each study

Study Criers (non- Physiology measured Cry-induction Method criers) Kraemer and Hastrup, 18 (16) HR; SCL 28-min emotional film (Peege) 1998 Marston et al., 1984 16 (5) N/A 55-min emotional film excerpts (The Champ) Delp & Sackenheim, Not reported Lacrimal flow 9-min total time remembering 1987 a sad personal event Labott et al., 1990 8 (8) S-IgA 28-min emotional film (Peege) Martin, Guthrie, & Pitts, 26 (16) S-IgA 18-min emotional news story 1993 Gross et al., 1994 33 (117) HR; SCL; RR; FT; 205-sec emotional film FPA; PTTF; PTTE; excerpt (Steel Magnolias) ACT Vingerhoets & 46* (19) Cortisol Emotional film- timing unclear Kirschbaum, 1997 (Shadowlands or Once Were Warriors) Rottenberg et al., 23 (104) HR; SCL; SCR; FT; 170-sec emotional film excerpt 2002** ACT; RR (The Champ) Sakuragi, Sugiyama, & 10 (0) HRV 50-min self-selected tragedy Takeuchi, 2002 scenes Hendriks et al., 2007 27 (33) HR; BP; SCL; RR; 70-min emotional film (Once RSA; PEP Were Warriors) Gračanin et al., 2015 28 (32) N/A 45 min emotional film excerpt (La vita è bella or Hachi: A dog’s Tale) Ioannau et al., 2016 13 (0) Facial Thermography 30-min (participant selected)

Note: *Study unpublished; **Study compared depressed to non-depressed participants with 17 depressed and 6 non-depressed criers overall; HR = heart rate; SCL = skin conductance level; S-IgA = secretory immunoglobulin A; RR = respiration rate; FT = finger temperature; FPA = finger pulse amplitude; PTTF = pulse transmission time to finger; PTTE = pulse transmission to the ear; S-Iga= secretory immunoglobulin A; ACT = general somatic activity; SCR= skin conductance response; HRV = heart rate variability; BP= blood pressure; RSA= respiratory sinus arrhythmia; PEP= pre-ejection period CHAPTER 3 60

Crying Theories

Lay people and researchers alike continue to champion the pervasive idea that the purpose of crying is to facilitate recovery and a return to homeostasis. Of the theories that have emerged to explain its functionality, crying is most often described as being part of a recovery process, with a ‘release’ or ‘overflow’ function that is utilized during extreme distress and is marked as the end of the distressing period (Efran & Spangler, 1979;

Vingerhoets, Bylsma, & Rottenberg, 2009).

Taking a reductionist view, Frey focused on the biochemical aspects of crying, in claiming that tears may help to detoxify the body during periods of distress that helps to improve mood (Frey & Langseth, 1985). Predominantly, he observed: (1) that distress produces biochemical changes within the body, (2) that people generally feel better after crying, and (3) that crying is an excretory process; that is, emotional tears caused by distress remove toxic substances from the body.

Panksepp (1998), on the other hand, theorized that crying may trigger the release of certain endogenous opioids, which may have an analgesic and pain reducing effect, rather than an exocrine process for which chemicals are leaving the body via tearing. Panksepp suggested that the primary function of these substances is to facilitate recovery after having been in distress. Perhaps these hypothesized effects may in some part explain the functional effect of crying. That is, crying could serve to sedate, reduce pain, restore the homeostatic balance, and even assist in coping with stressors as was suggested by Frey (Frey & Langseth,

1985; Vingerhoets et al., 2000). This theory presents thought-provoking ideas that have, thus far, not been directly empirically tested. Perhaps crying acts as a protective function to numb negative affective and physical pain, and to assist with recovery from distressing events or to increase coping resources. If so, this would help to explain why we cry when we are in pain, RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 61 and more significantly, why we cry when we are alone. This could also be seen as an explanation of why tearful crying continues into adulthood.

Current study

The current study aimed to examine whether crying promotes a reduction in the levels of stress hormones, namely cortisol, that may facilitate coping when experiencing physical and social stressors. This was investigated by randomly exposing participants to either a cry- eliciting stimulus or a neutral stimulus. However, as this was a quasi-experimental design, participants in the cry-eliciting condition were grouped as ‘sad-criers’ or ‘sad non-criers’.

From the reviewed literature, we hypothesized that emotional crying would facilitate recovery following experiences of distress such that crying would improve both physical stress responses and psychological mood, including improving a person’s ability to cope with future physical stress. Specifically, we predicted that compared to those who did not cry (both sad non-criers and those in the neutral group), participants who cried would: (1) rate worse mood, yet show the greatest improvement over time; (2) show lower heart rate and respiration during the video stimuli; (3) be able to cope with the stressful task for longer; and

(4) show lower physiological stress responses following the stressful task (cold pressor task).

Overall, given several laboratory experiments to date have shown worse mood following crying (Cornelius, 1997), we anticipated that participants would show this same pattern, with no change expected for those watching the neutral film. It was unclear what would happen following crying. However, along with a reduction in cortisol, and based on the analgesic and emotional recovery theories, we expected that participants who cried would feel less negative affect over time in comparison to the other groups and show lower cortisol levels overall. In addition, physiological measures of HR and respiration rate were predicted to diminish following crying. These changes were investigated among participants who were exposed to the cry-eliciting stimuli (both criers and non-criers) and for those who only CHAPTER 3 62 watched neutral stimuli.

Given the high number of participants needed, the time needed for data collection, and that pre-testing revealed a much greater proportion of women who cried to the stimuli

(73% compared to 29% of men), we included only women in the present study. Furthermore, due to methodological considerations for timing related to cortisol collection, self-report scales were included as a filler task between the collection of cortisol samples. Because these were not central to our original question, exploratory analyses were conducted on crying proneness, and other variables thought to be related to crying, such as empathy, emotion regulation, and social expectancies (Sharman et al., 2018). Information and results regarding these measures and results can be found in Appendix A. Rationale and key hypotheses for this study were specified before data collection, with complete data files and materials permanently archived at https://osf.io/z5uaw/.

Method

Participants

Participants were 197 female undergraduate students at the University of Queensland, each receiving credit toward an undergraduate psychology course. An additional twelve participants were excluded because either they were taking medications that may have interfered with physiological measures, they did not identify as female, or their testing involved experimental errors. Participants ages ranged from 17 to 50 years (M= 20.08, SD=

4.67) and they identified as 66% Caucasian/white, 22.8% Asian, 1% Pacific Islander, 0.5%

Aboriginal or Torres Strait Islander, and 9.1% other.

Stimuli

Videos. We used two different sets of film stimuli were used in this study. Both sets consisted of five short videos that were edited to play continuously and in the same order each time. The first set was used as a neutral stimulus, and these clips were chosen from RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 63 documentary scenes, Ted talks, and true stories. The combination of clips for the neutral stimulus set went for 17.5 min. The second set was used as a cry eliciting stimulus and included videos that induced crying in 46% of participants in a pilot study of both men and women. The video that was the highest rated for tearful responses in the pilot study was chosen to be the first video, with the other videos following to continue the sad emotion induction. This order was chosen to ensure the most tearful responses were captured at the beginning of video watching due to methodological restraints involving cortisol changes.

Specifically, as a change in cortisol requires time before it is detectable in saliva, the video that was related to the greatest number of cry-responses in pre-testing was presented first (see below in salivary cortisol). To ensure emotion induction continued during the period, all of the following videos were sad but did not necessarily elicit a cry-response in all participants.

The selected videos were from advertising campaigns, animated film scenes, and true stories.

Pre-testing found that there was no relationship to previous viewing of a video clip and whether a person cried or not in response to current viewing. The duration of the combination of clips for the sad stimulus set was 17.9 min.

Cold Pressor Stress Test (CPT). The CPT was used as the stressful induction and involved participants placing their left hand, up to the wrist, in cold water between 0-5°C.

This stimulus produces a slowly mounting pain of mild to moderate intensity and is terminated by voluntary withdrawal of the limb. The CPT has been used in many studies involving hormonal stress responses with generally moderate effects (Bullinger et al., 1984;

Geliebter, Carnell, & Gluck, 2013; Schwabe & Schächinger, 2018). The length of time

(measured in seconds) participants could keep their hand in the water was also recorded for each participant as a measure of coping ability, this was both manually recorded by an experimenter and checked using the video recording with any discrepancies corrected. CHAPTER 3 64

Measures

Heart rate and respiration. Heart rate was recorded using 10-mm pre-gelled

Ag/AgCl disposable electrodes attached over the lower rib on the left side of the torso and to the participant’s chest on both the right and left, according to published guidelines (Bernston

& Quigley, 2007), and recording a lead III electrocardiogram. Respiration was recorded using a respiration belt measuring abdominal expansion and contraction. HR, expressed as the number of beats per min (bpm), and respiration rate, represented as the number of breaths per minute, were sampled across the following time periods: final 1 min baseline, 1 minute of each video, and 15 seconds of CPT (See ‘Data Processing’ below regarding determination of times measured for each video segment). Data for both HR and respiration rate were collected together using a wireless connection to an MP150 BIOPAC ECG/RSP system, and the data were digitized at 200 Hz.

Salivary cortisol. Saliva samples were collected approximately every 10-20 mins throughout the study, yielding four samples. These were taken at 20 mins, 35 mins, and 45 mins from the initial sample. The second sample had a longer duration between samples because it was estimated that tears would be shed within at least 5 mins following the initial sample (i.e., when the first video began). Thus, an extra 15 mins were given following the initial crying episode to detect a change and allow for some variation in time for the first crying response. During collection, each participant kept a cotton swab in their mouth for 1.5 min, before returning the swab to a container. Saliva samples were stored in a −20°C freezer until study completion. They were then shipped to a professional reference laboratory at the

Technical University of Dresden, Germany, which assayed them for cortisol. Samples were centrifuged for 5 min, and hormone concentrations were measured by commercially available chemiluminescence-immuno-assays with high sensitivity (IBL International, Hamburg,

Germany). Intra- and interassay coefficients of variations were below 10%. RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 65

PANAS. Subjective emotional state was assessed with the Positive and Negative

Affect Scale short form (PANAS) at four separate time-points within the study using the time instructions “at the present moment” (Thompson, 2007). Participants were instructed to indicate the extent that they felt each of ten emotions rated on a 5-point scale from 1 (very slightly or not at all) to 5 (extremely). Five emotions were presented for both positive (e.g.,

“inspired” and “enthusiastic”), and negative (e.g., “afraid,” “upset,” and “ashamed”) valence.

Items were summed to create positive and negative affect scores at four time-points during the study. To remain consistent with previous research and to ensure positive emotions were not confounded with participants views that crying is something they may enjoy, despite the negative emotions they may be feeling, only the negative affect summary scores were analysed. Reliability for the Negative PANAS was found to be acceptable for all time-points

(α’s > .71), except the baseline, which showed a lower reliability of α = .55. As all other uses of the measure were acceptable, this was included within analyses.

Procedure

Participants were randomly assigned to the sad or neutral video condition at approximately a 60:40 sad: neutral ratio. This was decided on the basis of recent previous research with approximately 30% of people exposed to cry-eliciting stimuli induced to cry

(Gračanin et al., 2015). To avoid extraneous influences on HR and cortisol, participants were asked to refrain from smoking, eating, drinking caffeinated and alcoholic beverages for at least 2 hours before participating. Participants were also asked to refrain from vigorous exercise on the day of participation. This was confirmed upon arrival and any food or drinks consumed were discussed with the researchers. Participants were only allowed to continue if they met requirements to the researchers best judgement. Participants were screened for any contagious diseases and any medication they may be taking. Due to the effect of diurnal variation on cortisol, participants were only tested between 1 pm and 6 pm on any given day CHAPTER 3 66

(Follenius, Brandenberger, Hietter, Siméoni, & Reinhardt, 1982; Stone et al., 2001).

Information regarding the phase of the menstrual cycle was not recorded as no strong current evidence has found a link between specific phases and actual crying behaviour (Horsten et al.,

1997; Romans et al., 2017). Participants’ faces were also video recorded throughout testing to ensure crying could be subjectively rated and to ensure appropriate analyses were conducted

(see ‘analytic strategy’ below).

After attaching the electrodes, baseline heart and respiration activity were recorded while sitting silently for 5 min and participants were then asked to provide a saliva sample for baseline cortisol measurement directly following this. Participants were asked to complete the first set of PANAS questions. The video stimuli (sad or neutral) was then played to participants. Following watching, participants were asked to complete their affect ratings on the PANAS and rated their “most intense emotional experience” while watching the video

(1= no reaction, 5= choked up, 9= sobbing) using a cry self-rating scale. This single item measures self-ratings of crying and has been used previously in measuring crying responses to films (Labott et al., 1990). After these self-ratings, participants provided a second saliva sample. Participants then completed the CPT, another PANAS measure, and the other questionnaires, including the crying proneness scale (see Appendix A) and demographic questions. Approximately 10 minutes after the CPT, participants were asked to provide another saliva sample. To ensure participants did not leave the study feeling upset, and to gain a longer time frame of recovery from the videos and cold pressor test, participants watched a 10 min funny video before providing a final saliva sample, PANAS ratings and being debriefed. Ethics clearance was obtained through the University's School of

Psychology ethics committee. RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 67

Results

Data processing

All participants were evaluated for observable crying responses during video watching, to ensure that self-reports were aligned with a physical response. As previous research by Hill and Martin (1997) indicated the self-report measure to be highly correlated with behavioural observations of crying, assessment of crying was limited to whether crying did or did not occur in line with our research aims. That is, observed crying was not assessed based on more nuanced differences in crying that might be reflected in self-reports (e.g., lump in throat, tears in eyes, or tears down face).

A subset (half) of 34 participants who indicated they cried during video watching and were verified to have tears, were evaluated by two independent researchers who noted each

10-sec epoch of the videos determining when tears were present (including tears in eyes or down the face) or not during video-watching. These ratings determined 1-min segments of each video that were considered to be the ‘saddest’ with the highest number of criers2. These ratings determined clear peak emotional periods that caused crying and that were reliable across those who cried during each video. Crying frequency graphs for each of the sad videos can be found in supplementary materials. Inter-rater reliability was found to be 90%. To remain consistent in our comparisons, the same time segments were analysed in the neutral videos with a 20 second adjustment to account for the small overall difference in video duration. Measures for HR and respiration rate were then averaged over the agreed time- segments for each of the five videos, which then formed a single composite measure for video-watching.

2 Identified peak emotional response periods of each video involved: a dog with their family in a park as it is put to sleep (video 1); a woman finding out that her older sister had cancer (video 2); the death of a partner and their funeral (video 3); memory and animation of a phone call from a partner in one of the twin towers during September 11 (video 4); faces of people crying over family members lost to car accidents. CHAPTER 3 68

To ensure comparisons with baseline were also consistent in length, the final 1 minute of the initial baseline recording was used as the baseline measurement for both HR and respiration rate. For HR and respiration rate measures, only the initial 15 seconds were recorded based on previous findings that increases in HR during the CPT reaches its peak in the initial 16-26 seconds (Peng et al., 2015). Participants who could not withstand the CPT for at least 15 seconds were not included within these analyses.

SPSS version 24 (IBM, 2016) was used for all statistical analyses. Means and standard deviations (SDs) for all measures are reported in Table 3.2. Where sphericity assumptions were not met for the following results, Greenhouse-Geisser tests were reported.

Overall, participants were 71 in the sad-cry group, 61 in the sad-no-cry group, and 65 in the neutral group.

Self-report ratings of negative affect

Due to experimenter error or participants not appropriately completing scales, 11 participants were excluded from analyses of self-reported negative affect. Negative affect was analysed across four time points throughout the study in a 3(Group: Cry, No-Cry, Neutral) x

4(Time: 1 [baseline], 2 [post-videos], 3 [post-CPT], 4 [final]) mixed ANOVA, with Group as the between-subjects variable. These results revealed a significant main effect for Time

(F(2.45, 447.35)= 97.90, p < .001, η2p = .35), Group (F(2, 183)=13.97, p < .001, η2p = .13) and a Group x Time interaction (F(4.89, 447.35)= 28.93, p < .001, η2p= .24), as depicted in

Figure 3.1.

Tests of simple effects revealed significant effects for each Group; Cry: F(2.36, 94.91)

= 76.02, p < .001, η2p = .54; No-Cry: F(1.70, 94.91) = 44.53, p < .001, η2p = .44; Neutral:

F(2.18, 94.91) = 9.83, p < .001, η2p = .14. Simple comparisons revealed a similar pattern for the cry and no-cry groups with an increase in negative affect revealed from Time 1 to Time 2

(both ps < .001). A reduction in negative affect was found from Time 2 to Time 3 (both ps RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 69

Table 3.2: Means (SDs) and sample sizes for negative affect ratings, heart rate, respiration, and cortisol across each group at each time point in analyses performed

Group N T1-Baseline T2-Videos T3-Stressor T4-Final

Criers Heart Rate 59 84.21 (15.84) 81.71 (11.19) 91.73 (18.00)* 82.44 (10.09) Respiration 59 14.45 (4.08) 14.65 (2.88) 12.68 4.36)* 14.67 (3.76) Cortisol 70 5.51 (3.67) 4.65 (2.86) 4.23 (2.32) 3.96 (1.98) Negative affect 66 6.24 (1.66) 9.23 (2.79)* 6.58 (2.16) 5.26 (0.64)* Non-criers Heart Rate 47 79.74 (13.15) 81.49 (23.73) 85.96 (18.49)* 79.67 (15.24) Respiration 47 14.54 (3.73) 15.59 (2.16)* 12.03 (4.57) 15.34 (3.75) Cortisol 60 5.49 (3.20) 4.55 (2.53) 4.71 (3.75) 4.09 (2.51) Negative affect 57 6.35 (1.47) 8.40 (2.90)* 6.44 (2.04) 5.32 (1.10)* Neutral Heart Rate 46 79.31 (12.28) 76.96 (9.42) 85.27 (17.15)* 77.48 (10.73) Respiration 47 15.50 (2.76) 16.63 (2.16)* 13.23 (4.46) 15.37 (3.45) Cortisol 64 5.82 (4.39) 4.97 (4.19) 4.75 (3.93) 4.49 (4.33) Negative affect 63 6.14 (1.62) 5.71 (1.54)* 5.71 (1.54) 5.22 (1.08)*

NB: * Bonferroni corrected p < .05 difference compared to T1-baseline

10

Neutral 9 Cry No-Cry

8

* * 7

NEGATIVE NEGATIVE AFFECT 6

5

4 TIME 1 (BASELINE) TIME 2 (VIDEOS) TIME 3 (CPT) TIME 4 (FINAL)

Figure 3.1. Self-report ratings for negative affect for each group at baseline, after video watching, following the cold pressor test (CPT), and at the end of the study. Higher scores relate to worse mood. Error bars indicate standard error of the mean. *Signifies between group difference with Bonferroni correction p < .05. CHAPTER 3 70

< .001), with no differences found between baseline (Time 1) and Time 3 (both ps = 1.00).

Time 4 was significantly lower in negative affect than all other time points in both groups

(both ps < .001; see Figure 3.1). Simple comparisons for the neutral group revealed differences between Time 1 and 2 with a decrease in negative affect ratings, p < .001. No differences were observed from Time 2 to Time 3, p = .67. However, a decrease in negative affect was found at Time 4 compared to Time 1 (p < .001) and Time 3 (p = .03), but no difference was found between Time 4 and Time 2 (p = .25; see Figure 3.1).

Self-report ratings of negative affect

Due to experimenter error or participants not appropriately completing scales, 11 participants were excluded from analyses of self-reported negative affect. Negative affect was analysed across four time points throughout the study in a 3(Group: Cry, No-Cry, Neutral) x

4(Time: 1 [baseline], 2 [post-videos], 3 [post-CPT], 4 [final]) mixed ANOVA, with Group as the between-subjects variable. These results revealed a significant main effect for Time

(F(2.45, 447.35)= 97.90, p < .001, η2p = .35), Group (F(2, 183)=13.97, p < .001, η2p = .13) and a Group x Time interaction (F(4.89, 447.35)= 28.93, p < .001, η2p= .24), as depicted in

Figure 3.1.

Tests of simple effects revealed significant effects for each Group; Cry: F(2.36, 94.91)

= 76.02, p < .001, η2p = .54; No-Cry: F(1.70, 94.91) = 44.53, p < .001, η2p = .44; Neutral:

F(2.18, 94.91) = 9.83, p < .001, η2p = .14. Simple comparisons revealed a similar pattern for the cry and no-cry groups with an increase in negative mood revealed from Time 1 to Time 2

(both ps < .001). A reduction in negative mood was found from Time 2 to Time 3 (both ps

< .001), with no differences found between baseline (Time 1) and Time 3 (both ps = 1.00).

Time 4 was significantly lower in negative affect than all other time points in both groups

(both ps < .001; see Figure 3.1). Simple comparisons for the neutral group revealed differences between Time 1 and 2 with a decrease in negative mood ratings, p < .001. No RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 71 differences were observed from Time 2 to Time 3, p = .67. However, a decrease in negative mood was found at Time 4 compared to Time 1 (p < .001) and Time 3 (p = .03), but no difference was found between Time 4 and Time 2 (p = .25; see Figure 3.1).

Follow up comparisons with Bonferroni adjustments were also carried out between

Groups within each Time-point. These analyses found no differences between Groups at

Time 1, all ps = 1.00. However, a difference between groups was found at Time 2 following video watching, with the Neutral group showing the lowest negative affect compared to both criers and non-criers’ group (both ps < .001). No difference was found between the Cry and

No-cry groups at time 2 (p = .22); demonstrating that the sad emotion induction was effective regardless of whether the participant cried while watching. Time 3 displayed no differences for the Cry and No-cry groups, p = 1.00, or the Neutral and No-cry groups, p = .08, with only the neutral group showing significantly lower negative affect ratings than criers only, p = .02.

The final time point showed no differences between any groups, all ps = 1.00.

Cortisol

Given the time periods for collection of Cortisol were not identical, cortisol measurements across the four time-points of collection were combined to compute two measures: ‘area under the curve with respect to increase’ (AUCI) and the ‘area under the curve with respect to ground’ (AUCG) representing the total measure of hormone concentration to look at overall differences between Groups. These were computed following guidelines for cortisol analysis by Pruessner, Kirschbaum, Meinlschmid, and Hellhammer

(2003). To test hypotheses for individual time-points, cortisol results were compared in a series of one-way between groups ANOVAs. One participant’s data were removed due to elevated cortisol levels similar to cortisol awakening response with cortisol more than 4 SD above the mean at each time point. Two further participants were not included in analyses where cortisol data were missing at Time 4. CHAPTER 3 72

One-way ANOVAs examining both AUCI (F(2, 191) = 0.10, p = .90, η2 = .001) and

AUCG (F(2, 191) = 0.17, p = .85, η2 = .002) showed no effect overall of Group on cortisol responses indicating there were no differences in total hormone concentration depending on

Group (AUCI: Mneutral = -27.53, SD neutral = 86.43, Mcry = -32.54, SDcry = 62.32, Mno-cry =-28.26,

SDno-cry = -29.57; AUCG: Mneutral = 188.06, SD neutral = 138.51, Mcry = 176.76, SDcry = 98.93,

Mno-cry =180.24, SDno-cry = 103.64). Further analyses of each time-point for cortisol collection revealed no differences between Groups at any time-point; Time 1: F(2, 192) = 0.16, p = .86;

Time 2: F(2, 192) = 0.283, p = .75; Time 3: F(2, 191) = 0.49, p = .62; Time 4: F(2, 190) =

0.52, p = .59 (see Table 3.2 for means and SDs, and see Figure 3.2).

To further investigate these null effects, equivalence testing (Lakens, Scheel, &

Isager, 2018) was performed comparing criers to non-criers on the AUCi using the TOSTER module in jamovi (version 0.9; jamovi project, 2018). The TOST procedure for Welch’s t-test for independent samples, using equivalence bounds expected of ±0.50, revealed the effect of cortisol is statistically equivalent as both were found to be significant3, t(128) = 2.44, p

= .008. These results indicate that cortisol difference between criers and non-criers is smaller than originally predicted.

Cold Pressor Test

To examine whether crying helps participants to endure a physical stressor, all participants completed the cold pressor test. A one-way ANOVA revealed no differences between Groups on the amount of time spent attempting the CPT, F(2, 194) = 1.06, p = .35,

η2 = .01 (Mneutral = 46.88, SD neutral = 33.69, Mcry = 56.08, SDcry = 39.29, Mno-cry =50.59, SDno-cry

= 38.16). To investigate this null effect, equivalence testing was used to compare criers and non-criers on participants times for the CPT. Equivalence bounds were set to ±0.50 and both

3 Only the smaller test statistic of the two is reported when both are significant, as recommended by Lakens et al. (2018). RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 73

6.5

Neutral 6 Cry No-Cry

5.5

5 NMOL/L

4.5

4

3.5 TIME 1 (BASELINE) TIME 2 (VIDEOS) TIME 3 (CPT) TIME 4 (FINAL)

Figure 3.2. Mean cortisol levels for each group measured at baseline, following video watching, after the cold pressor test (CPT), and at the end of the study. Error bars indicate standard error of the mean. No significant differences found between groups.

tests revealed statistical equivalence, t(128) = 2.05, p = .02. Thus, the difference between criers and non-criers on endurance with a physical stressor is smaller than initially predicted.

Physiological Measures

Heart rate (HR) and respiratory rate data were analysed using a series of 3 (Group) X

4 (Time: Time 1 [baseline], Time 2 [average of videos], Time 3 [CPT], and Time 4 [Final minute] mixed ANOVAs, with Group as the between-subjects’ factor. However, thirty-six participant’s physiological (i.e., heart rate and respiration) data were lost due to computer failure. Of those remaining, participant data were rarely missing more than one data-point.

Therefore, to reduce the number of data deleted listwise in a four time-point repeated measures ANOVA, data were imputed when only one of four time-points were missing data.

This was imputed by including the average score for the time-point for the respective condition. Imputation for HR and respiration did not occur for a further seven participants CHAPTER 3 74 where participants did not withstand the cold pressor test for at least 15 seconds (see Data

Processing, above).

Heart rate. A total of 152 participants were included in HR analyses, with 44 participants excluded due to measurement problems at one or more of the four time-points.

Results found an effect of HR across time, F(3, 381.94) = 15.33, p < .001, η2p = .09, but no effect for Group or a Group X Time interaction. Pairwise comparisons of each time-point revealed no differences between Time 1, Time 2, and Time 4. However, differences were observed between Time 3 and the three other time-points (all ps < .001). This indicates that no differences were observed for any Group during the video watching period for HR compared to baseline. However, for each Group there was an increase in HR during the cold pressor test (see Figure 3.3).

Respiration. Analyses for respiration rate were conducted on 153 participants. A significant effect was found for respiration over Time, F(2.35, 352.74) = 22.37, p < .001, η2p

= .13, and for Group, F(2, 150) = 3.48, p = .03, η2p = .04 but no effect was observed for a

Group X Time interaction, see Figure 3.3. A follow up one-way ANOVA indicated there were no differences between groups at Time 1 (baseline), F(2, 156) = 1.04, p = 0.36.

Bonferroni adjusted pairwise comparisons of each time-point found differences between

Time 1 and Time 2 (p = .01), and Time 3 and all other time-points (ps < .001). These comparisons indicate that there was an increase in respiration rate during the video watching period following baseline, and a reduction, compared to baseline, during the cold pressor test, see Table 3.2 for means and standard deviations. Group level comparisons indicated a difference between Cry and Neutral groups, p = .03, with the neutral group showing higher respiration rates compared to those who cried. However, no differences were found between

Neutral and No-cry or Cry and No-cry groups. These results reveal that respiration rate was overall lower for those who cried, compared to those in the Neutral group during video RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 75

100

Neutral 95 Cry No-Cry 90

85

80

HEART HEART RATE (beats per min) 75

70 TIME 1 (BASELINE) TIME 2 (VIDEOS) TIME 3 (CPT) TIME 4 (FINAL MIN) 18

17

16 * 15

14

13

12

RESPIRATION RATE RESPIRATION RATE (breaths per min) 11

10 TIME 1 (BASELINE) TIME 2 (VIDEOS) TIME 3 (CPT) TIME 4 (FINAL MIN)

Figure 3.3. Mean heart and respiration rate measured at 4 time-points: at baseline, an average of the saddest points of five videos, throughout the cold pressor test (CPT), and during the final minute of the experiment. Error bars indicate standard error of the mean. *Signifies between group difference with Bonferroni correction p < .05.

CHAPTER 3 76 watching. Furthermore, although there was a reduction in respiration rate during the cold pressor test, there were no differences between groups, F(2, 150) = 0.85, p = .43, η2 = .01.

Further comparisons were carried out determining change during video watching time within each group. Paired t-tests showed that both the Neutral (t(48) = 3.07, p = .004, d = .50) and No-cry (t(50) = 2.14 p = .04, d = .32) groups showed an increase in respiration rate compared to baseline. However, there was no increase in respiration within those who cried during video watching (t(58) = -0.54, p = .59, d = .06).

Exploratory Results

Heart Rate and Respiration. In response to a reviewer’s suggestion, we conducted additional analyses to investigate the lack of findings for criers for both HR and respiration.

These additional analyses compared criers’ pre-crying and crying periods, similar to the methods used in Gross et al. (1994). A third rater assisted with coding the videos according to the initial crying onset. Following independent coding, the raters came together and re- watched each video based upon each rater’s coding and agreed upon time-segments for the beginning and end of the first crying onset for participants. This was determined by visible tears down the face or in eyes accompanied by changed facial expressions or face-touching indicative of crying. Raters ensured that these crying periods were also comparatively different to facial expressions and tearing before and after. Criers’ HR and respiration for crying periods were then extracted for 30 seconds of the cry period and 30 seconds before the cry period (pre-crying). For criers who showed shorter crying periods, their pre-crying and crying analyses were adjusted accordingly.

For HR, results showed no differences for criers between baseline (M = 84.19, SD =

15.86) and the first crying period (M = 82.96, SD = 12.15), t(55) = 0.70, p = .49, d = .09.

However, there was a significant reduction in HR during the pre-cry period (M = 79.87, SD =

12.36) compared to both baseline (t(56) = 2.81, p = .007, d = .37) and the first crying period RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 77

(t(55) = 2.99, p = .004, d = .40; see Figure 3.4). A somewhat different pattern was seen for respiration, in which no differences were found between baseline (M = 14.56, SD = 4.02) and the first crying period (M = 14.99, SD = 4.07), t(57) = 0.70, p = .49, d = .09, nor for the pre- cry period (M = 15.01, SD = 3.78) compared to baseline (t(57) = 0.78, p = .44, d = .10), or the first crying period (t(57) = 0.05, p = .96, d = .006).

Sensitivity Checks for CPT Endurance and Physiological Measures. Given the lack of group differences for CPT endurance, cortisol change, and HR, it was suggested that the binary coding of crying (crying vs. no crying) might not be sensitive to the emotional intensity related to crying. For example, those who did not cry may have almost cried, experiencing a lump in the throat instead, for example, and perhaps a continuum of emotional crying responses is related to physical responses. As such, binary coding may have masked these effects. However, it is essential to note that this may reveal effects of sadness, rather than effects of crying, specifically. To investigate this possibility, bivariate correlations were conducted to examine whether the self-reported intensity of emotional response to the videos was related to each of these measured physical responses, including for respiration rate, which did show effects at the group level.

Correlations across the three groups found no differences for self-reported emotional responses compared to time spent on the CPT (r = .12, p = .11) or cortisol change (AUCi; r =

-.09, p = .20). However, there was a relationship found for HR (r = .20, p = .02) and respiration rate (r = -.34, p < .001) during video watching. This indicates that greater emotional response to videos was related to increased HR and possible reduced respiration rate; however, the previous results for respiration rate reveal this was likely driven by an increase in respiration rate for non-criers, with no changes in those who actually cried, which remained at baseline.

CHAPTER 3 78

87

86

85

84

83

82 *

81

80 HEART HEART RATE (beats per min) 79

78

77 BASELINE PRE-CRY PERIOD FIRST CRYING PERIOD

Figure 3.4. Mean heart rate for criers measured three time-points: baseline, pre-crying, and initial crying periods. Error bars indicate standard error of the mean. *Signifies difference between time points p < .05.

In case there were differing responses for different video types, further correlational analyses were conducted for only those who watched the sad videos. These analyses found no relationships between sad video watchers self-reported emotional response and time spent on the CPT (r = 0.11, p = .24), cortisol change (AUCi; r = -.04, p = .62), or HR during video watching (r = 0.18, p = .07). However, respiration rate during video watching still showed a significant relationship, r = -0.25, p = .01.

Discussion

The present study aimed to establish whether crying reduces cortisol levels, and whether a reduction in cortisol may facilitate coping when experiencing stressors. To that end, female participants were assigned to a sad or neutral emotion induction, with 54% of women in the sad condition crying. This produced three groups: sad-criers (N=71), sad-non-criers (N=61), RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 79 and neutral (N=65). Ratings on the PANAS negative affect were the same across groups before the emotion induction (Time 1) and at the end of the experiment (Time 4), but negative affect was significantly higher after the emotion induction (Time 2) in the two sad groups than in the neutral group, regardless of whether the participant cried. Contrary to expectations, and despite this affective response to the sad emotion induction, those who cried were not able to cope with the stressful task for significantly longer, nor did they show any reductions in cortisol levels compared to those who did not cry.

Because crying is theorized to have such a powerful effect, our sample size was chosen based on a worst-case scenario-- a moderate effect. Small, non-significant, differences did appear between the groups with those who cried able to withstand the CPT for, on average, approximately 10 seconds longer than those in the neutral group and 6 seconds longer than those who did not cry to the sad stimuli. Given that the effect size for this result was quite small, if there is some regulation in the body involving chemical changes because of crying, it appears doubtful that it would prove useful in circumstances where you may require aid, and in particular, circumstances of physical injury. Equivalence testing for both cortisol and

CPT times found that true effects are more likely to fall below a medium effect size and would, thus, require very large sample sizes to find group differences. Consequently, the present data do not support theories on sad crying and the release of substances which sedate or reduce physical pain and improve coping to those physical stressors.

When interpreting these results, we note that cortisol showed no significant increase following the CPT, raising a question about the sensitivity of the measurement. Although we note that the cortisol measurements in this research followed a typical diurnal slope over time, suggesting that it was appropriately measured, the T3 measurement was followed by the CPT task for all participants. If we had included a no CPT-task group, perhaps we would have observed a steeper cortisol drop compared to those who did perform the CPT. It is also CHAPTER 3 80 possible that our CPT was not sufficient in this experimental context to elicit changes between criers and non-criers.

Heart Rate and Respiration

Our results found differences in respiration across groups. Notably, the data revealed that criers showed no change in respiration rate from baseline. However, participants in both of the non-crying groups displayed increases in respiration rate. It appears from these results that our method of emotion induction may have promoted increased respiration, with participants who cried seeming more capable of regulating their breathing and keeping it at a constant rate compared to those who did not cry but still viewed the sad videos, and also those who only watched neutral videos. Of particular interest was the stepped increase in respiration between each of the emotionally responsive groups, for no emotion, sad, and crying, which suggests that physiological change may not purely be reduced to whether emotional tears were present or not, but rather the intensity with which those emotions are felt and expressed. This result provides further support to the idea that crying assists the regulation of arousal.

Contrary to previous research, crying was not found to increase HR among criers either from baseline or in comparison to other non-crying participants (Hendriks et al., 2007;

Kraemer & Hastrup, 1988). Given the limited previous research measuring either HR or respiration rate and the potential lack of reliability due to small and unequal sample sizes, it is difficult to know if this result reflects a true lack of arousal during crying. Somewhat consistent with previous research, correlational results revealed that a greater intensity of emotional response across the three groups (neutral, non-criers, and criers) related to increased HR. Although this effect did not emerge when examining this same relationship among only those who were exposed to the cry-eliciting stimuli, as others have found (Gross et al., 1994; Hendriks et al., 2007). However, further analysis revealed that HR decelerated RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 81 directly before the crying period, comparable to results found by (Gross et al., 1994). This suggests that crying does have some effect on physiology and could lend support to either the arousal view of crying, that crying increases physiological arousal; or to the recovery view of crying, that crying helps to restore homeostasis. We conclude from our results, that it appears for both HR and respiration there is a relationship with homeostasis, such that crying appears to return HR to homeostasis, and may be involved in the regulation of breathing to keep respiration at a stable rate. It is also possible that these two processes work in tandem with either conscious or unconscious attempts to regulate breathing during sadness, which may in turn influence changes in heart rate.

Crying and Emotion

Crying is generally regarded as the result of an emotional peak, especially that of sadness or helplessness, but this is also extended to many other discrete emotions including happiness, anger, and frustration (Miceli & Castelfranchi, 2003; Vingerhoets et al., 2009).

However, given the difficulty in finding this relationship of sadness with emotional tears as more intense than that of sadness without emotional tears, perhaps crying does not indicate an increase in emotional intensity. Maybe tearing is primarily a social signalling cue indicating a need for help, with those that witness this response more likely to provide support and comfort, potentially strengthening social bonds (Gračanin et al., 2018). Nonetheless, the social signalling hypothesis does not support the paradoxical nature of crying alone or feeling better when we do (Rottenberg, Bylsma, & Vingerhoets, 2008). Perhaps crying is also an intrapersonal cue that signals a need to engage in regulation strategies to assist with managing distress, rather than a peak of recovery and release (Efran & Spangler, 1979;

Vingerhoets et al., 2001, 2009). Or perhaps, this personal signalling also engages the crier in alternative strategies to relieve or manage distress such as actively seeking others who may provide support or assistance, rather than crying acting solely to cue the helper. Either of CHAPTER 3 82 these speculations may help to explain why people report feeling better after crying, yet it has not been observable in most laboratory inductions of crying (but see: Gračanin et al., 2015).

However, of course, a person’s attitudes about whether crying is helpful and a positive distress reduction strategy is likely to play a mediating role in the actual emotional effects

(Sharman et al., 2018), and we do not know if these findings generalize to coping with social stress for which crying could serve a different function.

Although in the broader research context, researchers (including us) primarily focus on negative affective crying - particularly crying caused by sadness - it is noteworthy that laughter often occurs with the production of tears. Laughter, too, is often prescribed as having significant mental and physical health benefits (R. A. Martin, 2012; Provine, 2012). However, the additional element of tears produced when laughing are not generally thought to have any positive effects on their own. The production of tears in joyful contexts are often thought of as by-products of the experienced emotion, which by and large do not alter the emotional experience, but are generally interpreted as a heightened emotional experience. Some have speculated that these tears are similar to reactionary tears produced when we yawn, for example, and are potentially caused by the strong contraction of the muscles surrounding the eyes that produce tears (Gračanin et al., 2018). Interestingly, unlike crying from sadness, which is widely regarded to be beneficial with little direct evidence, laughter has been found to have analgesic properties, by increasing endorphins and endogenous opioids (Dunbar et al.,

2012; Manninen et al., 2017). It is curious that crying from sadness, on the other hand, has long been thought of as being highly connected with the tears that are produced and that they are often believed to have an extremely important role in both the emotional experience and its recovery from negative feeling states. This has not necessarily been the case when people consider tears of joy. RESPONSES TO STRESS FOLLOWING TEARS OF SADNESS 83

Conclusion

Results from this study overall provided no support for the notion that crying promotes a change in biochemical substances that support our physical coping ability. Specifically, no effects were found for changes in cortisol, or time spent on the CPT for criers and non-criers, whether they watched the cry-eliciting stimuli or not. However, compared to those who did not cry and whose respiration rate increased, respiration appeared consistent for those who cried, with crying, perhaps, aiding the regulation of breathing for these participants.

Furthermore, heart rate showed a return to homeostasis after a reduction directly before the crying period.

Concerning these results, however, we also acknowledge the limitations in this research met by data failure of some physiological recordings that may have impacted our results. A further and considerable limitation of this study, and many laboratory studies like it, is the artificiality with which crying was induced in participants, and without any personal connection to the crying per se. We may have seen differences in crying and other measured responses among participants with more naturalistic changes in context and elicitation procedures, such as crying induced at a person’s home through the recall of personal, highly emotional events; this may be a more ecologically valid way to study crying in future research.

From our results, we speculate that crying may promote self-soothing and regulation by initiating emotion regulatory behaviours in criers and, more so, that this response may be more linked to the intensity with which an emotion is experienced. However, we also found no differences between criers and non-criers in their experiences of sadness following video watching, and this seems to be a trend among some crying research. Thus, we further consider that crying may also be an intrapersonal cue signalling necessary engagement in emotion regulation strategies or the seeking of others for assistance in order to reduce distress CHAPTER 3 84

(Gračanin et al., 2018). The felt improvement that some feel following crying may indeed be the result of the physical homeostasis we retain by crying, as well as the social support that we receive during or after crying, and the increased mental coping efforts we engage in to help us feel better when experiencing a negative emotion (and which do not necessarily have any relationship with crying). Finally, perhaps popular notions about the benefits of crying are perpetuated by our personal and social evaluations of crying as helpful or useful, and not necessarily whether these physical homeostatic mechanisms provide actual physical stress- relieving benefits. THE BELIEFS ABOUT CRYING SCALE 85

CHAPTER 4: Does Crying Help? Development of the Beliefs About Crying Scale

This chapter is from a manuscript that has been published in Cognition and Emotion:

Sharman, L. S., Dingle. G. A., & Vanman, E. J. (2019). Does Crying Help? Development of

the Beliefs About Crying Scale (BACS). Cognition and Emotion, 33, 722–736.

https://doi.org/10.1080/02699931.2018.1488243

Contribution to authorship:

The conception and design of the project was completed by me (70%), Dr Vanman (15%), and Dr Dingle (15%). All data was collected and analysed by me. Selection of appropriate items before factor analysis was assisted by both co-authors. Important theoretical interpretations and decisions regarding all factor analysis were made by me. The entire manuscript was drafted and edited (including professional revision decisions) by me with critical review provided by both co-authors. CHAPTER 4 86

Abstract

Crying is often considered to be a positive experience that benefits the crier, yet there is little empirical evidence to support this. Indeed, it seems that people hold a range of appraisals about their crying, and these are likely to influence the effects of crying on their emotional state. This paper reports on the development and psychometric validation of the Beliefs about

Crying Scale (BACS), a new measure assessing beliefs about whether crying leads to positive or negative emotional outcomes in individual and interpersonal contexts. Using 40 preliminary items drawn from a qualitative study, an exploratory factor analysis with 202 participants (50% female; aged 18-84 years) yielded three subscales: Helpful Beliefs,

Unhelpful-Individual Beliefs, and Unhelpful-Social Beliefs, explaining 60% of the variance in the data. Confirmatory factor analysis on the 14-item scale with 210 participants (71% female; aged 17-48 years) showed a good fit to the three factors. The subscales showed differential relationships with measures of personality traits, crying proneness, emotion regulation and expressivity, and emotional identification (alexithymia). The BACS provides a nuanced understanding of beliefs about crying in different contexts and helps to explain why crying behaviour may not always represent positive emotion regulation for the crier.

THE BELIEFS ABOUT CRYING SCALE 87

Among many animal species, infant crying serves as a powerful caregiving signal.

Surprisingly, however, humans are currently believed to be the only species that continue to cry into adulthood (Vingerhoets, 2013). For adults, crying tends not to occur in instances of physical pain or hunger, as is often true for human infants. Rather, adult crying occurs when experiencing significant emotional pain or joy. In particular, adult crying has predominantly been interpreted as a type of cathartic release to help a person overcome negative emotions and ultimately- to increase positive emotions. For example, Efran and Spangler (1979) argued that “all psychologically induced tears signify release from tension”. Indeed, media outlets and scientific publications have almost uniformly promoted the positive value of crying and seem to have changed little in that view over time. A review of 72 articles published in popular media over 140 years found that 94% suggest that crying is beneficial to wellbeing (Cornelius, 2001). Unsurprisingly, many of these articles also emphasized the negative impacts that suppressing crying may have on personal wellbeing, without consideration of the diverse circumstances in which crying may occur, and which may make crying problematic or distressing for the crier. Yet, no consistent scientific evidence for the notion that ‘crying is good for you’ has been found (Vingerhoets, 2013).

Theories on the psychological functionality of crying also overwhelmingly promote the benefits of crying. Various psychoanalytical, cognitive, and social views on the functions of crying speculate that crying is a way of soothing and maintaining balance by reducing internal or external tensions, signalling help, and acting as a type of ‘safety valve’ for regulation of extreme emotion (Breuer & Freud, 1955; Efran & Spangler, 1979; Frijda, 1986;

Kottler, 1996; Labott & Martin, 1988; Lögfren, 1966; Sachs, 1973). Fewer theories have addressed the negative consequences of crying. Some, however, propose that crying is a sign of helplessness and powerlessness, leading to an increase in negative emotions felt by the crier (Frijda, 1986; Miceli & Castelfranchi, 2003). In particular, these theories identify that CHAPTER 4 88 crying occurs when a person feels that something is impossible to solve or that they are incapable of undoing/fixing the situation (Miceli & Castelfranchi, 2003).

Feelings of helplessness may also be amplified in social contexts. Crying in the presence of others can often make people feel judged by others as weak, incompetent, or emotionally unstable (MacArthur & Shields, 2015). A crier may feel further helplessness to prevent this negative social judgement, which may increase other negative feelings of worry, shame, and/or stress (Simons et al., 2013). Despite this, the strongest empirical evidence for the functionality of crying points to the positive social aspects of crying. Specifically, crying aids individual recovery by attracting others to provide help and succour when in distress

(Balsters et al., 2013; Hendriks et al., 2008; Vingerhoets, 2013). For the general population, however, potential associated stigma and beliefs that crying in interpersonal contexts may negatively affect them are more often associated with the social crying experience, rather than help and comfort seeking (Simons et al., 2013).

Previous research on emotional experiences in general has found that personal expectations and beliefs about an emotional experience may actually direct the emotional experience (Bastian et al., 2012; Eid & Diener, 2001). Similarly, the regulation of crying appears to be driven by the crier’s motivation either to increase feelings of personal catharsis, or to reduce negative social consequences by inhibiting their crying (Simons et al., 2013;

Vingerhoets et al., 2000). In particular, attempts to down-regulate (i.e., reduce or stop) crying have been associated with interpersonal motives to manage the negative social consequences associated with crying, such as being pitied or ridiculed. On the other hand, up-regulation of crying has been linked most often with individual motivations, such as achieving catharsis

(Simons et al., 2013). Although this evidence does point to the motivations for crying and techniques used in emotion regulation, little is known about how evaluations of crying may modify a person’s expression of crying and their accompanying appraisal of mood following THE BELIEFS ABOUT CRYING SCALE 89 it.

Currently, only two studies have explored both positive and negative evaluations of crying. The first was a large, cross-cultural, self-report study spanning 30 countries (Becht &

Vingerhoets, 2002). Participants reported their ‘general’ mood following crying compared to before, as well as their experiences of shame while crying, and an estimated crying frequency for the last month. Despite most participants reporting improved mood following crying episodes in general, mood change after crying was greatly affected by social and cultural factors at play for the individual. For example, negative mood changes in men were associated with greater shame, whereas mood change for women was associated with gender empowerment. Specifically, positive mood changes for women after crying, were associated with greater participation of women in political and economic decision-making within their country. Interestingly, masculinity-femininity, as well as national income of the participants’ country, were important variables influencing crying frequency and mood changes after crying for both men and women. Notably, however, other research on gender and culture has found that crying frequency continues to show a stronger relationship with biological factors rather than country characteristics of gender empowerment (Fischer et al., 2004). That is, crying frequency is still higher among women regardless of women’s status and social roles within a country. Nevertheless, these findings suggest that individual perceptions and beliefs about crying are related to the interpretation and appraisal of mood change after crying but may not be related to actual crying frequency.

The second study investigated mood change after crying among women as well as individual difference measures of personality and clinical functioning (Rottenberg, Bylsma,

Wolvin, et al., 2008). These results indicated that neuroticism, extraversion, and empathy predicted differences in frequency and ease of crying. Interestingly, none of these factors predicted mood change after crying. However, associations were found with worse mood CHAPTER 4 90 following crying and higher scores on alexithymia, depression, and anxiety. This indicates that some stable person factors may be less related to post-crying mood, whereas factors directly related to emotional experience, such as alexithymia, anxiety, and depression may determine how people perceive their crying and consequently, the influence of crying on their mood. For example, a person with alexithymia may feel confused and uncomfortable when crying because of an inability to describe and identify the emotions they are experiencing and may be more likely to interpret crying as a negative experience.

It is clear from the literature that crying involves an appraisal process that depends on idiosyncratic characteristics of the crier. Crying has been found to be related to a person’s age and gender, as well as individual difference variables, such as personality factors (Peter,

Vingerhoets, & van Heck, 2001), coping styles, and general proneness to crying (Denckla et al., 2014; Labott & Teleha, 1996; Vingerhoets, van den Berg, Kortekaas, van Heck, & Croon,

1993; Williams, 1982). These characteristics may affect whether a person engages in crying at all, but equally, how that person evaluates processes that may be occurring during crying.

Even though several person factors associated with crying have been identified, little consideration has been given to personal beliefs about crying. Similar to expectations of general emotional experiences (Bastian et al., 2012), the frequency with which a person engages in crying, how they feel after crying, and whether crying helps them to cope with an emotional event, are also likely to be influenced by their beliefs and expectations about crying, the social context, and past experience. Specifically, if a person who is experiencing a high level of distress cries and then feels some relief, he or she might believe that crying is positive and may be more likely to engage in crying. In contrast, a person who experiences no relief, or who feels helpless and weak after crying, might believe that crying is negative or unhelpful and may avoid crying. Of course, evaluations that crying is useful to help one overcome distress, may also alternatively encourage a person to cry so they can make THE BELIEFS ABOUT CRYING SCALE 91 themselves feel better, rather than making an evaluation post-crying. Thus, beliefs are likely formed because of a combination of prior experience (e.g., feeling better after crying) and relevant social expectancies (e.g., crying should make one feel better, but it will also be judged negatively in many social situations).

To date, research on beliefs about crying in various contexts has been hampered by the lack of a validated measure. Despite a relatively robust literature supporting the positive effects of crying in social contexts, and somewhat unclear research regarding the mood reduction effects of crying (Bylsma et al., 2011; Gross et al., 1994; Rottenberg, Bylsma,

Wolvin, et al., 2008; Vingerhoets, 2013), people are more likely to use crying to make themselves feel better and to reduce negative social consequences. It appears then that our beliefs about crying do not necessarily map onto the empirical space, and this may help to explain why crying research has not always been consistent. A more in-depth understanding of what individual beliefs about crying are would be useful in both research and clinical areas to understand how and why crying is used in different contexts for different people, what personal factors may be related to different beliefs, and whether these beliefs intersect with other psychological constructs (see Table 4.1).

This study

Different appraisals of crying are expected to lead to different emotional effects after crying, and to different levels of coping following an emotional event. This study reports on the development and psychometric validation of a measure of crying beliefs that assesses the beliefs that people have about their crying behaviour and whether crying leads to positive or negative outcomes for the individual in various contexts. This research was conducted in four phases. First, beliefs about crying items were generated using a qualitative study of 28 adults.

Next, the set of 40 items were subjected to an exploratory factor analysis with a sample of

202 adults. Third, the measure was validated using confirmatory factor analysis with a new CHAPTER 4 92 sample of 210 adults. Finally, the new subscales were tested for criterion and discriminant validity alongside existing measures of crying and emotion regulation. Below we report how we determined our sample size, all data exclusions and all measures used in this study. The complete data files and questions used are permanently archived at https://osf.io/22y4y/?view_only=17c5fddfa1ff4f2a808ad5724d020c12.

Construction of the BACS

Item Generation

Participants and Procedure. To ensure an adequate range of qualitative responses for item generation, we decided that a minimum of 25 participants would be satisfactory.

Overall, 28 participants (17 male; 11 female) aged between 19 and 57 were recruited online

(M = 30.65, SD = 8.53) through the crowdsourcing platform Prolific

(https://www.prolific.ac). From this sample, 77% identified as White, 11% African/African-

American, 8% Asian, and 4% Hispanic/Latino. Participants were asked to respond to questions regarding how they believed crying would affect them using common antecedents of crying outlined in previous research (Miceli & Castelfranchi,

2003). Questions were open-ended and related to their beliefs about crying in general or in specific contexts related to previous crying literature. See Appendix B for questions used.

Participants were also asked to complete the Kessler Psychological Distress Scale (K10;

Kessler et al., 2002) to ensure interpretation of crying beliefs were not related to overall distress levels due to some evidence of crying items on depression inventories correlating with depression severity. Because we did not know if this would affect evaluations of crying, extreme scores on this scale were assessed for responses that differ substantially from other participant responses.

THE BELIEFS ABOUT CRYING SCALE 93

Table 4.1: Psychological Variables Potentially Related to Beliefs about Crying

Variables Hypothesised Relationship People who believe that crying is helpful may be more prone to crying, especially if Crying proneness they think crying will make them feel better.

Emotion self-regulation difficulties may be related to beliefs that crying results in

Emotion regulation poorer outcomes if a person is not able to effectively manage distressing emotions that would likely be related to crying events.

If it is difficult to know or understand what emotions a person is feeling, then

Understanding emotions difficulties in understanding emotions would be related to poorer beliefs about crying. Crying, in particular, may result in greater confusion and misunderstanding.

Neuroticism is the personality trait most consistently found in crying research, with findings that neurotic people cry more easily and frequently for a variety of reasons,

Personality and given its associations with negative mood states and reactions (De Fruyt, 1997; Mathell et al., 2001; Muris et al., 2005), it is likely that neurotic people would interpret crying as a negative experience.

Women would be more likely to believe crying is positive, given that are more Gender likely to perceive crying as a coping strategy (De Fruyt, 1997; Timmers et al., 1998).

Item Selection

Answers provided by participants were reviewed and analysed for individual beliefs about crying and explored for common themes among all answers. Because men were overrepresented in our sample, individual responses for men and women were compared to assure there were no marked differences between genders in responses. In fact, there was considerable overlap in content and theme, with both genders reporting that crying either helped to increase or decrease negative feelings, or made no overall difference to how they felt in all situations presented. This included mentions from both men and women of using crying for distraction from physical pain, experiencing either guilt or comfort in social situations, feelings of worthlessness when experiencing failure, and cathartic release of anger.

Only one response (provided by a woman) was not mirrored in theme by any male respondents, which was a comment about the potential to use crying to manipulate others CHAPTER 4 94 when responding to a question about crying when you are with others.

As this area of research was largely unexplored, items were drawn from both common and single mentions of beliefs, especially if uncommon responses (e.g., crying to manipulate others) were present in the crying literature. Positive or negative versions of items were included to ensure questions addressing different moods were included. After item creation and modifying items for clarity, this process created an initial pool of 40 items (see Appendix

B for full list of items). These items represented a breadth of crying beliefs in a range of relevant contexts that have previously been identified in the crying literature. No items were excluded based on scores from the K10.

Exploratory Factor Analysis

Participants

An exploratory factor analysis was conducted on a second volunteer sample of 202 participants, (sample determined by a minimum of 5 participants per item; Comrey & Lee,

1992) recruited online through Prolific (https://www.prolific.ac). This sample included 101 men and 101 women, with ages ranging from 18 to 84 (M = 35.29, SD = 13.77). Participants identified as 84% White, Asian 8%, African/African-American 5%, Hispanic/Latino 2%, and

Other 1%.

Statistical Approach and Results

SPSS version 24 (IBM, 2016) was used for the exploratory factor analysis. Frequency examinations of all items showed no high intercorrelations (r > .90) between items. One item with excessive skew (+/- 2; “crying helps me to get what I want”) was revealed during this process and removed before analysis. To determine the extraction method, the remaining items were assessed for multivariate normality, with skewness (Ŷ1p = 389.65, p = 8.07) and kurtosis estimates (Ŷ1p = 1687.25, p = 11.09) indicating non-normality (Korkmaz, Goksuluk,

& Zararsiz, 2014). To account for the distribution of the sample, we used Principal Axis THE BELIEFS ABOUT CRYING SCALE 95

Factoring (PAF).

The factorability of the 39 items was assessed before analysis with the Kaiser-Myer-

Olkin (KMO) measure verifying sampling adequacy for analysis as good, KMO = .84. All

KMO values for individual items were > .6, above the acceptable limit of .5 (Kaiser, 1974).

Bartlett’s test of sphericity, 휒2 (741) = 4321.53, p < .001, indicated that correlations were sufficiently large for our analyses. We used PAF with a Promax (oblique) rotation to allow for correlations among the factors in the exploratory factor analysis. To ensure the scale sufficiently reflected important aspects of the measure, the factor solution was based on a combination of the scree test, Kaiser-Guttman rule (factors with an Eigenvalue > 1), and clear factor loadings.

A nine-factor solution was initially revealed with 57% of variance explained, all factors displaying eigenvalues greater than 1, and accounting for the most variance. However, few items displayed primary loadings outside of three factors. A three-factor solution was ultimately retained as the simplest factor structure, with eigenvalues of 8.52, 6.69, and 3.10, respectively, and accounted for 42% of the total variance. Items were then removed if they displayed poor parameter estimates (items with target loadings < .50) or cross-loadings > .20.

We examined the communality statistics to ensure the retained items accounted for the greatest variance in the data. Items with communalities < .40 were examined closely, with two items ultimately removed because they did not closely relate to their respective factors

(Costello & Osborne, 2005). With the removal of these items, the total explained variance increased. Finally, the item “crying makes me feel immediately better” was removed for redundancy with several other items (e.g., “crying makes me feel better”, and “I feel less stressed after I cry”, and “after crying I feel an emotional release”). These steps ensured the items reflected the strongest aspects of the measure with the largest total variance explained.

Overall, a three-factor structure was retained with 14 items that explained 60% of CHAPTER 4 96 total variance (factor loadings for each item are presented in Table 4.2). The first factor comprised of seven items reflecting beliefs that crying aids emotion regulation when the person is alone (helpful beliefs). The second factor comprised of four items reflecting beliefs that crying makes one feel worse when around others (unhelpful social beliefs). The third factor included three items reflecting beliefs that crying makes one feel worse when alone (unhelpful individual beliefs).

Means, standard deviations, correlations, and Cronbach’s coefficient alphas are reported in Table 3. Raw score item correlations within each scale were between .36 and

.74. The subscales were moderately correlated, except no correlation was found between helpful and unhelpful-social. These results suggest that each scale is distinct, with beliefs associated with crying in the presence of others being especially different from beliefs about crying when alone. This means that it is possible for a person to hold both strong positive beliefs about crying as helpful to them on an intrapersonal level while also endorsing beliefs that crying negatively impacts them when surrounded by others. Alternatively, a person with low scores on both the helpful and unhelpful-individual scales is more likely to indicate that crying ‘does nothing’ and produces no emotional change for them. Because of this, we recommend that each of the subscales be interpreted separately, rather than combined as a total score in analyses.

Due to the general differences in crying frequency of men and women, and with women more likely to perceive crying as a coping strategy (Becht & Vingerhoets, 2002; De

Fruyt, 1997; van Tilburg et al., 2002), independent sample t-tests were also conducted on each subscale to determine any gender differences that might be present. Analyses revealed differences between men and women only on helpful beliefs, t(200) = 2.84, p = .005, d =

.39, with women more likely to endorse beliefs that crying is helpful for emotion regulation.

However, the mean difference was small at 0.37 with a small effect size. Since these gender THE BELIEFS ABOUT CRYING SCALE 97 differences were small and expected, no further analyses were conducted as a function of

Table 4.2: Factor Loadings of the Beliefs About Crying Scale

No. Items Helpful Unhelpful- Unhelpful- Social Individual 14. Crying makes me feel better. 0.85 11. Crying helps me to process my emotions. 0.83 -0.103 9. In the long run, I know that I'll feel better because I have cried. 0.807 10. I feel less stressed after I have cried. 0.793 -0.112 1. After crying I feel an emotional release. 0.723 7. Crying helps when I'm feeling overwhelmed. 0.702 0.14 2. Crying helps me to express how I'm feeling without words. 0.602 0.118 5. I feel ashamed when I cry around people who are not my friends or family members. 0.933 8. I feel judged when I cry around co-workers. 0.788 12. It's embarrassing when I cry around friends or family. 0.74 6. Crying around others makes me feel vulnerable. 0.678 4. Crying makes me feel worse when I'm alone. 0.128 0.835 13. In the long run, I feel worse because I have cried. 0.757 3. I feel worse immediately after I cry. 0.675 Note: Factor loadings reflect values from the pattern matrix. Bolded coefficients denote primary factor loadings.

gender.

The exploratory factor analysis revealed a 14-item scale distributed over three factors.

These factors represent beliefs that crying makes one feel better (helpful); that crying makes one feel worse when alone (unhelpful individual); and that crying makes one feel worse when others are present (unhelpful social). Overall, the subscales demonstrated good internal consistency, with α’s between .8 and .9 (see Table 3). Small gender differences on the helpful beliefs subscale are consistent with previous research reporting gender differences in crying CHAPTER 4 98 proneness and frequency between men and women (Becht & Vingerhoets, 2002; De Fruyt,

1997; Timmers, Fischer, & Manstead, 1998; van Tilburg et al., 2002).

Confirmatory Factor Analysis and Discriminant and Convergent Validity

Participants

The three-factor structure of the measure was validated with a confirmatory factor analysis using a third sample of 210 participants (60 male; 150 female). Sample size was again determined with a minimum of five participants per item (Comrey & Lee, 1992).

Participants completed the 40-item version of the scale along with several measures of convergent validity. Participants were undergraduate students who received course credit for participation. They were aged between 17 and 48 years (M = 20.18, SD = 4.79), and identified as 62% White, 31% Asian, 0.5% Aboriginal and Torres-Strait Islander, and 6.7%

Other.

Criterion Measures

To check for convergent and discriminant validity, the following measures were chosen because they represented constructs that were either similar or conceptually different to crying beliefs (Campbell & Fiske, 1959). All scales were completed by the third sample used for the confirmatory factor analysis. Participants completed these scales along with the

40-item version of the BACS scale. For each of the chosen scales our aim was to understand how beliefs about crying may be associated with proneness to crying; emotion regulation strategies both inter- and intra- personally; whether differences arise between understanding and communication of emotions and individual beliefs; and how personality dimensions may be associated with each of the positive, negative, and negative-interpersonal beliefs4.

Predictions regarding these constructs can be found in Table 1. To understand the association

4 The Jackson-5 was also measured as an alternative personality measurement. However, as some research has suggested significant limitations in the measurement of r-BIS in the Jackson-5, these results were not analysed (Corr, 2016). THE BELIEFS ABOUT CRYING SCALE 99 emotion regulation strategies may have with beliefs about crying, both personal (Difficulties in Emotion Regulation Scale) and social (Interpersonal Emotion Regulation Questionnaire) emotion regulation scales were chosen to reflect the scales on the BACS. Given previous associations found between crying, personality, and clinical functioning (Rottenberg, Bylsma,

Wolvin, et al., 2008), alexithymia, an inability to identify and describe emotions, was measured using the Toronto Alexithymia Scale, and personality traits were assessed using the

Big Five Inventory. The crying proneness scale was also assessed as the most recent and well-validated crying scale to measure associations with crying proneness and crying beliefs.

Crying Proneness Scale (CPS; Denckla et al., 2014). The CPS consists of a 25-item

7-point scale (1= very unlikely, 7 = very likely) assessing crying proneness in various contexts and with differing precipitants (Denckla et al., 2014). The scale consists of four sub-scales assessing crying for attachment, sentiment, society, and compassion. All scales displayed good internal consistency within our sample with reliability for all scales between .77 and

.95.

Difficulties in Emotion Regulation Scale short form (DERS; Gratz & Roemer,

2004). Intrapersonal emotion regulation was measured through the DERS. This is an 18-item scale measured on a 5-point scale (1 = almost never, 5 = almost always) with questions focusing on problems with emotion regulation. The short form of this scale has previously shown excellent internal consistency across the six subscales: (1) strategies (e.g., “When I’m upset it takes me a long time to feel better”), (2) non-acceptance (e.g., “When I’m upset I become embarrassed for feeling that way”), (3) impulse (e.g., “When I’m upset I become out of control”), (4) goals (e.g., “When I’m upset I have difficulty concentrating”), (5) awareness

(e.g., “I care about what I am feeling”), and (6) clarity (e.g., “I am confused about how I feel”), with our sample showed similar reliability between .73 and .91 across all subscales. CHAPTER 4 100

Interpersonal Emotion Regulation Questionnaire (IERQ; Hofmann, Carpenter,

& Curtiss, 2016). The Interpersonal emotion regulation questionnaire is a new scale consisting of 20 items with four sub-scales: (1) Enhancing positive affect (e.g., “I like being around others when I’m excited to share my joy”), (2) perspective taking (e.g., “having people telling me not to worry can calm me down when I am anxious”), (3) soothing (e.g., “I look to others for comfort when I feel upset”), and (4) social modelling (e.g., “It makes me feel better to learn how others dealt with their emotions”). Overall, the scale shows excellent psychometric characteristics with reliability for all scales between .81 and .91 in our sample.

Emotional Expressivity Scale (EES; Kring, Smith, & Neale, 1994). General emotional expressivity was measured through the EES. The EES is a 17-item scale measured on a 6-point scale (1= never true, 6 = always true) showing excellent internal reliability across multiple samples (alpha coefficients ranging from .90 to .93) and also within our sample, α = .93. Items on this scale reflect a person’s perceptions of their expressivity (e.g.,

“I can’t hide the way I’m feeling”) and how they believe others perceive their emotional expressions (e.g., “other people believe me to be very emotional” and “the way I feel is different from how others think I feel”).

Toronto Alexithymia Scale (TAS-20; Bagby, Taylor, & Parker, 1994).

Alexithymia (the inability to identify emotions) was assessed using the 20-item TAS-20 measured on a 5-point scale (1 = strongly disagree, 5 = strongly agree). Within our sample, reliability for the total TAS-20 was found to be good at .85.

Big Five Inventory (BFI; John & Srivastava, 1999). Personality measures were taken using the BFI assessing five different personality traits of openness, conscientiousness, extraversion, agreeableness, and neuroticism over 44-items on a scale from 1 = disagree strongly to 5 = agree strongly. This scale has a strong factor structure and has shown acceptable internal consistency across subscales both generally and in online samples of THE BELIEFS ABOUT CRYING SCALE 101 between .79 and .88 (John & Srivastava, 1999; Srivastava, John, Gosling, & Potter, 2003).

This was consistent in our sample with internal consistency ranging between .75 and .86.

Results

Confirmatory Factor Analysis. Confirmatory factor analyses were conducted in R version 3.3.3 (R Core Team, 2017), package extension lavaan .5 to 23.1097 (Rosseel, 2012).

Multiple indices were used to evaluate model fit using established cut-offs: the root mean squared error of approximation (RMSEA), standardized root mean residual (SRMR), comparative fit index (CFI), and non-normative fit index (Tucker-Lewis Index; TLI; Hooper,

Coughlan, & Mullen, 2008; Hu & Bentler, 1999). Acceptable fit was determined by RMSEA and SRMR < .07 and < 0.08 respectively, and CFI and TLI > .90 (Hu & Bentler, 1999;

Steiger, 2007).

To evaluate the possibility of a simpler structure comparing only positively and negatively valenced items, a two-factor solution was also compared to the three-factor solution in a confirmatory factor analysis to establish which solution gave the best fit to the data. The two-factor model included positive beliefs in one factor and the negative items in a second factor (negative individual and interpersonal beliefs). Maximum likelihood estimation was used with both robust and non-robust estimates applied. As there was little difference between the two, only robust estimates are reported here. Using the fit indices outlined previously, it was determined that the two-factor model fit the data poorly 풳2(76) =

222.691, p < .001, CFI = 0.88, TLI = 0.85, RMSEA = 0.096, SRMR = 0.12. However, as predicted, fit for the three-factor model proved substantially better, 풳2(74) = 159.483, p <

.001, CFI = 0.94, TLI = 0.93, RMSEA = 0.067, SRMR = 0.072.

Thus, the confirmatory factor analysis supported the three-factor solution produced by the exploratory factor analysis. The means, standard deviations, inter-correlations and alphas for each of the three factors are reported in Table 4.3. These scores show few differences to CHAPTER 4 102 the previous sample reported in the EFA, and only showed a slight reduction in reliability coefficients for the individual negative beliefs scale, but given the scale has a small number of items we still considered this to be acceptable at >.60.

Independent sample t-tests were conducted on each scale to determine any gender differences within this sample. Similar findings emerged as for the previous sample, in that differences between men and women were only found on the positive individual subscale t(208) = 3.98, p < .001, d = .59; with women more likely to endorse beliefs that crying is helpful for emotion regulation. The mean difference was slightly larger than the previous sample at 0.56 with a moderate effect size.

Convergent and discriminant validity. Correlations were conducted between the

BACS and each of the scales, with correlation values shown in Table 4.4. Strong negative correlations were found between helpful beliefs and the crying proneness scale overall with small to no associations with both unhelpful scales. Although this result seems unintuitive, items in the crying proneness scale are related to factors that influence crying external of the individual, for example, ‘how likely is it that you are touched to the point of tears when you see/read/hear … a wedding ceremony’. In contrast, the helpful beliefs scale contains items that focus on the person’s experience in a situation, rather than their reaction to an external event, for example, ‘I believe I feel safe when I cry alone’. This correlation suggests then, that if people believe that crying is helpful for their personal emotion processing, they are less prone to crying in contexts that are not self-centred/involved.

Positive relationships were found between each of the difficulties in emotion regulation sub-scales with either one or both of the unhelpful beliefs scales, suggesting that a person’s belief that crying is unhelpful both individually and socially is related to greater difficulties in emotion regulation. However, positive correlations were also found between helpful crying and the subscales for goals, impulse, and a negative relationship with THE BELIEFS ABOUT CRYING SCALE 103 awareness. Indeed, beliefs that crying is helpful were also related to feeling less in control of behaviours and difficulty focusing on tasks. Despite this, it seems that people who hold

Table 4.3: Scale Means, Standard Deviations, Inter-correlations, and Cronbach’s Alpha for

Individual Scale Reliability for the EFA and CFA Samples

EFA Sample CFA Sample Scale 1. 2. 3. 1. 2. 3. 1. Helpful 2. Unhelpful-Individual -.162* -.316** 3. Unhelpful-Social .073 .293** -.100 .296** Scale M 3.04 2.05 3.60 3.48 2.06 3.54 Scale SD 0.94 0.87 1.13 0.96 .78 1.01 Cronbach’s coefficient (α) .90 .80 .87 .91 .61 .80

Note: For correlations, *p < .05, **p < .001.

beliefs that crying is helpful also show greater awareness of their internal emotion states. For the interpersonal emotion regulation questionnaire, all sub-scales showed positive associations with helpful beliefs, indicating that seeking others when experiencing both positive and negative emotions is associated with greater endorsement of beliefs that crying is helpful. Only the ‘soothing’ subscale on the IERQ showed a negative correlation with unhelpful-social beliefs, intuitively suggesting that seeking others to help soothe one’s emotions was less likely to occur for those who scored highly on the unhelpful-social BACS.

Alexithymia, measured by the TAS-20, revealed relationships with both the beliefs of crying increasing interpersonal beliefs and general worse mood in both unhelpful scales. This is consistent with previous research indicating that alexithymia was associated with worse mood post-crying (Rottenberg et al., 2008). A positive relationship on helpful beliefs and negative relationship on unhelpful-social beliefs scales were also found with the EES. These results suggest that people are more likely to generally express emotions if they believe CHAPTER 4 104 crying is helpful to them and less likely to express emotions overall if they believe crying in the presence of others will lead to feelings of shame and judgement.

Table 4.4: Correlations between BACS and other self-report measures for the third sample

Scale Helpful Unhelpful- Unhelpful- Individual Social Demographics Age .765 .223 .002 Gender .266** .023 .057

TAS-20 Total -0.058 .261** .292**

EES .363** -0.115 -.375** IERQ

Enhancing positive affect .293** 0.009 0.051

Perspective Taking .193** -0.006 -0.076

Soothing .396** 0.039 -.141*

Social modelling .342** 0.052 -0.049 DERS

Strategies 0.001 .310** .276**

Non-acceptance 0.041 .356** .408**

Goals .251** 0.118 .344**

Impulse .209** .171* 0.067

Awareness -.320** .140* .215**

Clarity 0.057 .222** .160* CPS

Total -.367** 0.046 0.126

Attachment -.372** 0.059 0.123

CPS Societal -.253** -0.022 0.124

CPS Sentiment -.226** 0.041 .155*

CPS Compassion -.404** 0.048 0.052 BFI

Extraversion 0.055 -0.013 -.354**

Agreeableness .215** -.158* -0.044

Conscientiousness 0.048 -0.075 -.186**

Neuroticism .222** .172* .279** Openness 0.009 -0.004 0.007 Note: TAS-20 Total = Toronto Alexithymia Scale, EES = Emotion Expressivity Scale, IERQ = Interpersonal Emotion Regulation Questionnaire, DERS = Difficulties in Emotion Regulation Scale, CPS = Crying Proneness Scale, BFI = Big Five Inventory; *p< .05, ** p< .01 THE BELIEFS ABOUT CRYING SCALE 105

Associations between the BACS and the BFI showed negative relationships with unhelpful-social beliefs on both extraversion and conscientiousness. This suggests higher scores on the unhelpful-social beliefs scale is related to lower extraversion (introversion) and lower conscientiousness (impulsivity). Agreeableness was positively related to helpful beliefs and negatively related to unhelpful beliefs. Lastly, neuroticism was positively related to each of the three subscales on the BACS.

General Discussion

Because personal beliefs about crying can vary greatly between people and these evaluations may affect the likelihood of crying, as well as the emotions experienced after crying, we developed a new measure, the BACS, to assess those beliefs. Here we reported how across three studies with different samples, we chose the items, their factor structure, and their convergent and divergent validity.

Psychometric Properties of the BACS

The psychometric properties of the BACS were good overall and supported a three- factor structure with distinct subscales: helpful beliefs, unhelpful-individual beliefs, and unhelpful-social beliefs. Notably, the factor analyses supported the separation of negative crying beliefs into social and individual beliefs rather than combining these two scales into an overarching ‘negative beliefs’ scale. This was further validated by the differential correlations between these negative BACS subscales and measures of emotion expressivity, using others for emotion regulation, and personality dimensions. Specifically, higher scores on the unhelpful-social scale were related to lower endorsement of looking to others to help with emotion regulation, less emotion expressivity in general, and lower scores on extraversion and conscientiousness. Scores on the unhelpful-individual scale, however, were CHAPTER 4 106 not related to any of these measures. Despite their correlation, this pattern of results supports the notion that the two negative BACS subscales tap different constructs.

Surprisingly, despite the strong empirical support for the functionality of crying in social situations to gain help and succour (Balsters et al., 2013; Hendriks et al., 2008;

Vingerhoets, 2013), no ‘positive-social’ beliefs scale was revealed. We posit that this is likely because the significant negative stigma associated with crying in front of others (e.g., as embarrassing or weak) is stronger than any potential positive associations. This three-factor structure ignoring positive-social beliefs provides some evidence that any positive-social beliefs from the criers point of view are not strong, or not collectively held. Moreover, these results highlight problems in previous literature assessing the effectiveness of crying in social contexts almost exclusively from the point of view of the ‘social other’ and not of the individual who is crying. Unfortunately, our beliefs about what is helpful, does not appear to match the known literature in this area and may need to be reviewed in future research seeking to understanding whether the presence of others is actually helpful, or if it hinders recovery.

Gender

Gender differences were revealed on the helpful subscale. Specifically, women were more likely to endorse beliefs that crying was helpful to them for emotional regulation, with no differences found on either of the other scales. This provides further evidence that women, more than men, endorse that crying is ‘good’ for them, and this may be influenced by gender- stereotypical norms or expectations (Bronstein et al., 1996; De Fruyt, 1997; Fischer &

Manstead, 2000; Timmers et al., 1998). Perhaps this effect stems from a gender bias portraying crying among girls and women as acceptable, and crying from boys and men as weak (MacArthur & Shields, 2015). It was also revealed that beliefs that crying makes a person feel worse both individually and socially displayed no gender effects. We suggest that THE BELIEFS ABOUT CRYING SCALE 107 this may be because negative feelings about crying have had little discourse in Western societies, with the primary focus having been on how good crying is for improving wellbeing, with no emphasis on the possible negative impact of crying (Cornelius, 2001).

Indeed, Western cultures tend to place a higher value on positive, rather than negative, emotions, with this devaluation of negative emotions not always apparent in other cultures

(Bastian et al., 2012; Diener & Diener, 1995). This may be of particular interest for future cross-cultural work, and crying research in general, to explore the negative impact effect of crying on people and how this may differ depending on cultural expectancies of crying.

Crying, Emotion Regulation, and Emotion Expression

Beyond adequate psychometric qualities, the BACS displayed good convergent validity on scales related to emotion regulation and emotional understanding. In particular, those who believed that crying was good for their wellbeing and overall helpful, were more aware of their emotions, more emotionally expressive, and were also more likely to use others to help them with emotion regulation. Conversely, those who believed that crying was unhelpful for them either when crying alone or in the presence of others showed greater difficulties with emotion regulation, including less awareness of their emotions, and greater scores on alexithymia. This indicates that people who believe that crying makes them feel worse both in inter and intra-personal contexts, are not only less conscious of their emotions, they appear to be less able to understand or communicate their feelings.

Our findings relate to the broader literature on emotion expression, regulation and expectations of emotion experience in both personal and social contexts. Emotion dysregulation especially has been linked to the maintenance of psychological distress in general and particularly to those with alexithymia (Compare, Zarbo, Shonin, van Gordon, &

Marconi, 2014; Verissimo, Mota-Cardoso, & Taylor, 1998). People who are alexithymic, in particular, have been found to engage in greater maladaptive behaviours to regulate emotions, CHAPTER 4 108 such as binge eating and alcohol abuse, compared to those without alexithymia who were more likely to engage in constructive activities (Schaffer, 1993; Wheeler, Greiner, &

Boulton, 2005). This may explain why those who were scored higher on both unhelpful scales displayed greater emotion dysregulation than those who believed that crying was helpful for improving emotions. Furthermore, findings on the unhelpful-social scale appear closely linked with research on social norms and expectations about emotions (Bastian et al.,

2012; Campbell-Sills, Barlow, Brown, & Hofmann, 2006). It is possible that those who perceive crying to be unacceptable or believe that others expect them to feel positive, are more likely to suppress their emotions in an effort to reduce outward expression of that emotion. Because of this, they are more likely to feel worse following crying than those who feel their emotions are acceptable and congruent with how others think they should feel.

Crying and Personality

Convergent and divergent validity was also assessed between the BACS and personality dimensions on the BFI. Neuroticism, agreeableness, and extraversion have previously been associated with frequency of crying and were also associated with subscales on the BACS. In particular, more agreeableness was related to beliefs that crying is helpful when alone. Given that agreeableness has previously been associated with crying frequency, it is possible such people may be crying because they believe that crying is helpful and may engage in crying to make themselves feel better.

Our results also revealed that less extraversion and conscientiousness were associated with beliefs that crying in social contexts is unhelpful. Lower levels of extraversion have been associated with less involvement in social situations, a reluctance for attention, and social anxiety (Stein & Stein, 2008), so crying in a social situation likely amplifies uncomfortable feelings and may therefore reduce beliefs that crying in that context will lead to a positive outcome. Lower levels of conscientiousness, on the other hand, are associated THE BELIEFS ABOUT CRYING SCALE 109 with unplanned reactions to stimuli, or impulsivity. In particular, impulsivity has been related to non-constructive communication, and greater avoidance-withholding communication (Tan,

Jarnecke, & South, 2017). We speculate that because of this impulsive nature, people may regard emotional social situations negatively because of poor communication strategies, which may lead to a general belief that emotional exchanges, including crying, in these settings are unhelpful.

Neuroticism, surprisingly, was related to all three subscales. We speculate that perhaps neuroticism may direct strategies used to manage distressing experiences. That is, if crying is believed to be positive and helpful, then the emotional experience preceding may not be interpreted as negative. Indeed, neuroticism has previously been linked to perceptions of crying as a useful coping strategy (De Fruyt, 1997), similar to the content of the items on the helpful beliefs scale. However, neurotic people who believe crying is unhelpful may ruminate on the preceding negative experience and feel worse, consistent with previous literature in this area (Hervas & Vazquez, 2011; Muris, Roelofs, Rassin, Franken, & Mayer,

2005; Roberts, Gilboa, & Gotlib, 1998).

The Question of Causality

An anonymous reviewer questioned our assumptions regarding the causal direction between beliefs about crying and crying behaviour. That is, the development of this scale was grounded in the assumption that one’s beliefs about crying determine one’s crying behaviour.

However, as the reviewer suggested, this assumption precludes consideration that the beliefs may arise to explain or justify crying behaviour. For example, someone who cries frequently may come to believe it has a positive effect to justify why he or she cries so often, whereas someone who does not cry as frequently may in turn come to believe that crying has no effects. Although we do not have specific data to speak to this issue, an analysis of the relationship between the BACS and crying proneness scale showed that greater crying CHAPTER 4 110 proneness was related to fewer helpful beliefs about crying, and that less proneness to crying was related to more beliefs that crying is helpful. Although these results do not show support for the potential reverse causal pathway, it is still an important question to investigate in the future using methods designed to assess dynamic relationships between cognitions and behaviour. It may very well be, for instance, that beliefs about crying are updated throughout the lifespan as one experiences different social and intrapersonal outcomes from crying.

These updated beliefs then could influence future crying until a time when the beliefs are no longer functional and need to be updated again.

Conclusion

This research provided an opportunity to systematically evaluate different crying beliefs and determine their relationship to other individual factors. Although there were limitations for gender balance in the first sample for item creation, there were no apparent systematic differences in responses between genders, with both men and women responding with the same themes. More problematic is the limited diversity in each of the samples, with each predominantly representing a White and Western population. Given the diversity of adult crying, this scale should be tested cross-culturally to ensure it is appropriate for non- western samples. Further work should also be done to test the robustness of these findings, especially in relation to individual difference factors related to the BACS. Understanding the relationship between personality dimensions and the BACS may be a particularly interesting avenue for future research that may help to further identify the independent contributions in the variance of adult crying.

Future investigations might also consider the usefulness of the BACS, not only in general crying research, but also as a therapeutic tool. In particular, crying in therapy has been linked to negative evaluations of the therapy session, with client crying related to evaluations that a session was difficult and negative (Capps, Fiori, Mullin, & Hilsenroth, THE BELIEFS ABOUT CRYING SCALE 111

2015). These evaluations appear to increase in negativity the more crying occurs. Positively, however, this does not seem to affect the client-therapist relationship. The use of this measure in therapy may provide a clearer picture of how a client may evaluate crying. It may assist therapists in not only being aware of a client who may be adversely evaluating constructive therapy sessions, but also as an aid in understanding when and for whom it may be beneficial to explain and emphasise the safety of space in therapy and that crying can be a helpful way of emotional understanding and expression.

CHAPTER 5 112

CHAPTER 5: The Relationship of Gender Roles and Beliefs to Crying in an International

Sample

This chapter is from a manuscript that has been published in Frontiers in Psychology, section

Emotion Science:

Sharman, L.S., Dingle. G.A, Baker, M., Fischer., A.H, Gračanin, A. Kardum, I., Manley, H.,

Manokara, K., Pattara-angkoon, S., Vingerhoets, A.J.J.M., & Vanman, E.J. (2019).

The relationship of gender roles and beliefs to crying in an international sample.

Frontiers in Psychology, section Emotion Science.

https://doi.org/10.3389/fpsyg.2019.02288

Contribution to authorship:

The conception and design of the project was completed by me (75%), Dr Vanman (15%), and Dr Dingle (10%). Translations and data collection for each country were completed by me, Dr Vanman and Dr Dingle for Australia, as well as online data collection for some of the

Dutch and United Kingdom samples; Dr Vingerhoets, Dr Fischer, and Mr Manokara for the

Netherlands; Dr Gračanin and Dr Kardum for Croatia; Dr Manley and Ms Pattara-angkoon for Thailand; and Mr Baker for the United Kingdom. All data were analysed by me. The entire manuscript was drafted and edited by me with review provided by all co-authors. CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 113

Abstract

This study aimed to (1) investigate the variation in self ascription to gender roles and attitudes toward gender roles across countries and its associations with crying behaviours, emotion change, and beliefs about crying, and (2) understand how the presence of others affects our evaluations of emotion following crying. This was a large international survey design study (N = 893) conducted in Australia, Croatia, the Netherlands, Thailand, and the

United Kingdom. Analyses revealed that, across countries, genderm self ascribed gender roles, and gender role attitudes were related to behavioural crying responses, but not related to emotion change following crying. How a person evaluates crying, instead, appeared to be highly related to one’s beliefs about the helpfulness of crying, irrespective of gender. Results regarding crying when others were present showed that people are more likely both to cry and to feel that they received help around a person that they know, compared to a stranger.

Furthermore, closeness to persons present during crying did not affect whether help was provided. When a crier reported that they were helped, they also tended to report feeling better following crying than those who cried around others but did not receive help. Few cross-country differences emerged, suggesting that a person’s responses to crying are quite consistent among the countries investigated here, with regard to its relationship with a person’s gender role, crying beliefs, and reactions to the presence of others.

CHAPTER 5 114

The way in which people remember and evaluate their crying experiences are influenced by many factors. These may be whether the cause of crying was positive or negative, how long since the crying occurred (with crying remembered more positively the more time has passed), and social variables, such as the number of people present (Bylsma et al., 2011; Rottenberg et al., 2008; Vingerhoets, 2013). When researchers consider the social effects of crying, they primarily focus on specific social contexts. However, often neglected in these considerations are other social determinants, such as the socialisation of gender roles, which likely play a part in how humans engage in and evaluate their crying experiences.

Althouth it is well-established that women cry more than men (see Vingerhoets, 2013) this difference does not innately appear between sexes, with no such differences appearing among infants and small children (see Vingerhoets & Scheirs, 2000). From the age of 11, however, differences in crying proneness and frequency begin to emerge (van Tilburg et al.,

2002). Of course, contrary to the notion that women might be biologically inclined to be more emotional, research investigating the role of menarche and crying have found no association with menstrual cycles and crying behaviours (Romans et al., 2017; van Tilburg et al., 2003). Further, much of this change appears to stem from a reduction in boys’ expressions of crying, rather than an increase in girls’ (van Tilburg et al., 2002). However, this still leaves us with the finding that women tend to cry more, report more proneness to crying, and often feel better after crying than their male counterparts (Becht & Vingerhoets,

2002; De Fruyt, 1997; Denckla et al., 2014; Peter et al., 2001; van Hemert et al., 2011;

Vingerhoets & Scheirs, 2000).

The Role of Culture

Although cross-cultural research has been limited within the crying literature, some research has found a consistent gender difference across at least 35 countries (Becht &

Vingerhoets, 2002; Fischer et al., 2004). That is, women cry more frequently and tend to CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 115 report more positive crying experiences across cultures. This difference is particularly pronounced in many Western countries, with women showing considerably greater frequencies of crying as compared to women in Asian, South American, and in some West and East African countries (Becht & Vingerhoets, 2002). Despite these gender differences in frequency, however, in some countries, the difference in emotion improvement after crying between men and women was considerably smaller, and in some instances, non-existent, with gender overall explaining very little variance in emotion change.

Further analysis of this research has found that socio-cultural factors may play a role in how crying behaviours are expressed and evaluated (Becht & Vingerhoets, 2002). Indeed, in countries that were more gender equal, wealthier, and where people reported crying often and with little feelings of shame, both men and women tended to report feeling better after crying than people in countries with less equality and wealth. For crying behaviours, the more individualistic a country, the greater the reported crying frequency was for both men and women, with greater gender empowerment related to higher crying frequency among women. However, research by Fischer et al. (2004) alternately found that country level gender empowerment showed no relationship to crying frequency and concluded that crying behaviour is more strongly determined by biological rather than social factors.

These results suggest that the influence of culture on gendered differences in crying may play a role in influencing crying behaviours and evaluations. However, research in this area has primarily examined these potential influences using country level indexes and constructs of different socio-cultural factors (e.g., country gender empowerment). They were not able to investigate other possible individual difference factors related to gender and emotion expression. Here, we discuss two related avenues that may help to further explain gender differences found in crying research: (1) individual variation in the acceptance of gender and emotion norms (Fischer & Lafrance, 2015; Frymier, Klopf, & Ishii, 1990; Zahn- CHAPTER 5 116

Waxler, Friedman, & Cole, 1996), and (2) individual differences in crying beliefs (Sharman et al., 2018) that may affect engagement in crying and how it is evaluated.

Gender and Emotion Expression Norms

The widespread distinction between masculinity and femininity in the expression of negative emotions has resulted in a gender-role-consistent pattern of emotion expression, with men tending to express more powerful and hostile emotions such as ‘anger’ and having less tolerance for emotions that display vulnerability, such as sadness and shame (MacArthur

& Shields, 2015; Zeman & Garber, 1996). Women, on the other hand, tend to express more vulnerable emotions, such as sadness, fear, and shame, rather than anger (Fischer &

Manstead, 2000; Fischer & Lafrance, 2015). Indeed, early parental socialisation of emotion expressions in children appears to encourage more expression of anger and less expression of sadness in boys, whereas the opposite is true for girls, with parental socialization resulting in more displays of sadness and less anger than boys from as early as pre-school (Brody, 2001;

Chaplin, Cole, & Zahn-Waxler, 2005; Fivush, Brotman, Buckner, & Goodman, 2000).

Furthermore, there is an important influence of peers on emotion expression, whereby children are more likely to emphasise gender-role-consistent emotion behaviours when interacting with their peers (Chaplin & Aldao, 2013; Zeman & Garber, 1996). In cultures where gender differentiation for emotion expression is pronounced, this socialisation tends to culminate in the attitude that showing vulnerable emotions is a weakness for boys, illustrated with the well-known saying that ‘boys don’t cry’ (MacArthur & Shields, 2015).

Given this socialisation, perhaps unsurprisingly, differences in shame felt when crying have been found, with men reporting greater shame than women across the countries studied (Becht & Vingerhoets, 2002). However, this effect was small, and indeed, gender differences, particularly regarding mood and emotion ratings of crying across some of crying research have shown small or no effects (e.g., Lombardo, Cretser, Lombardo, & Mathis, CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 117

1983; Peter et al., 2001; Williams, 1982). Other research has alternatively focused on individual differences in beliefs about social roles. For example, Ross and Mirowsky (1984), employing a US sample, focused on men’s crying to understand if sex-role orientation showed an association with high or low crying expressions, with the authors hypothesizing that non-traditional men were more likely to reject stereotypical masculinity. Analyses revealed a relationship between the strength of men’s sex-role orientation and their crying frequency, such that men who adhered to more traditional sex-roles reported lower crying frequencies in response to sadness than non-traditional men. These results have led to suggestions that the differences found between men and women in crying research might be better explained by a person’s perception of gender role patterns (Vingerhoets et al., 2000).

That is, the extent to which a person endorses attitudes and behaves consistently with their gender role influence both their behaviour and emotion evaluations of crying. Despite these claims, however, there is a dearth of research on gender roles and their influence on crying behaviours and evaluations. Although there are clear gendered differences in reported crying behaviours and emotional evaluations following crying, understanding the influence of individual differences in gender roles may help researchers to understand some of the variability in crying experiences reported. Moreover, the measurement of crying behaviours has often been limited to crying frequency, which provides an estimate of recent crying but does not provide more detailed information regarding a single crying experience.

Crying Beliefs

Crying research has tended to focus on how people evaluate their crying experiences by measuring whether people feel better or worse following crying (see Becht &

Vingerhoets, 2002; Bylsma et al., 2011). Alongside monitoring changes to their own internal mood and emotional states across time, people also appear to use broader evaluative CHAPTER 5 118 strategies to understand their crying experiences through beliefs that they hold about how crying works generally and in different social contexts (Sharman et al., 2018).

Crying beliefs likely develop through a combination of prior experience and social expectations regarding crying. Indeed, general beliefs about the benefits of crying have changed little over the last 150 years with crying in the media almost consistently promoted as beneficial and, if suppressed, harmful to wellbeing (Cornelius, 2001). Yet, despite this overall saturation in the media, women are still more likely to endorse beliefs that crying is positive. Certainly, there is some evidence that beliefs about the effect of crying show small gender differences, with women believing that crying is more helpful than do men (Sharman et al., 2018). Although the research on crying beliefs is limited, these positive beliefs might help to explain why women rate feeling better than men following crying, where expectations and beliefs about an emotional experience might help to direct that emotional experience

(Bastian et al., 2012). However, these gender differences were not always consistent, with beliefs that crying was unhelpful in private or social situations showing no gendered differences (Sharman et al., 2018). Given that men are more likely to be socialised to believe that crying is a weakness, particularly in social situations, we would expect beliefs to show more of a gendered effect. These results suggest that beliefs about crying may not be as salient to gender stereotypes and because of this, these beliefs may actually mediate the impact of gendered emotion norms surrounding crying.

The current study investigated the individual variations in gender roles across genders and its links with crying behaviours, emotion change, and beliefs about crying. Our primary aim was to understand the function of gender roles and socialization in crying beliefs, crying intensity, and change to emotion following crying, regardless of country. Importantly, in this research we have referred to emotion, rather than mood as others have done (e.g., Becht &

Vingerhoets, 2002). Specifically, we were interested in negative emotions of relatively short CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 119 duration that caused crying and how those emotion states may have changed immediately following crying, rather than moods, which are generally considered to be longer lasting

(Rosenberg, 1998). Further, we also distinguish between gender role attitudes and self- ascription to gender roles. As crying behaviours, beliefs, and emotional evaluations likely encompass an interplay of both attitudes toward gender roles (e.g., “crying is for girls”) and degrees to which someone personally subscribes to those gender roles, we include both measures in this research.

We hypothesised that there would be relationships between reported gender, gender role attiutdes, self-ascribed gender roles, beliefs that crying is helpful, change in emotion following crying, and crying intensity. Such that (H1a), women would be more likely to endorse more feminine gender roles, with men more likely to endorse more masculine gender roles. Further, (H1b) participants who endorsed more masculine gender roles would report less intense crying, (H1c) worse emotions following crying, and (H1d) beliefs that crying is less helpful overall. Conversely, participants who endorsed more feminine gender roles would display higher crying intensity, improved emotion ratings following crying, and beliefs that crying is more helpful overall.

We hypothesized that (H2) the relationship between gender roles and perceived emotional change following crying would be accounted for by beliefs that crying is helpful.

However, we expected that effect to be observed when recency of the crying episode was accounted for, as crying is more likely to be remembered as generally more beneficial the more time that has passed since a crying episode (Vingerhoets, 2013). Thus, a stronger endorsement of feminine gender roles would be related to more beliefs that crying is helpful, leading to more positive emotion ratings following crying. We similarly hypothesized (H3) that the relationship between gender endorsement and crying behaviour (i.e., intensity of crying) would be mediated by beliefs that crying is helpful. Specifically, participants who CHAPTER 5 120

Figure 5.1. Model of predicted pathways between gender roles, beliefs about crying, and crying intensity.

subscribed to more feminine gender were expected to report beliefs that crying is helpful, leading to greater crying intensity (i.e., time spent crying and the strength of tearing during the last episode5; see Figure 5.1). Notably, despite general cross-country differences we believed that gender roles and their respective relationships with crying would remain consistent across countries. To explore whether these relationships were sensitive to country differences, mediation models (H2 and H3) were analysed to check if this effect is robust across countries, or whether these relationships differed depending on country.

Although the primary aims of this research address the potential influence of gender roles in crying behaviours and emotion change following crying, we were also interested in the extent to which the social context might influence people’s evaluations of emotion change following crying. That is, what role does the presence of others play in determining whether a crying person feels better, worse, or experiences no change after crying?

Crying and the Social Context

Evolutionary explanations of crying have suggested that crying mainly serves interpersonal purposes, such as soliciting help and facilitating social bonds (Gračanin et al.,

2018). Crying research has, thus far, shown convincing evidence for this view, finding that

5 As we were specifically interested in a person’s reports from their last crying experience, we did not include the general measure of crying frequency as a behavioural measure in the predicted mediation model CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 121 shedding tears in the presence of others appears functional. Specifically, the evidence indicates that crying is likely a social signal that encourages more helping behaviours, sympathy, succour, and less avoidance from nearby others (Hendriks et al., 2008; Hendriks &

Vingerhoets, 2006; Provine et al., 2009). However, overwhelmingly the methodology used in this research has involved participants rating their willingness to help when presented with vignettes or images of crying (or sad) faces, with little exploration from the perspective of the crier. Moreover, the available evidence that crying encourages helping behaviours in others comes from experimental studies conducted in primarily Western contexts.

Some research focussed on the crier’s experience has begun to emerge, with research comparing people who had lost the ability to cry to those who could. For example, Hesdorffer,

Vingerhoets, and Trimble (2017) found no differences between the two groups on their ratings of well-being but did find that people who could cry felt more socially connected with others than those who could not. These findings suggest that crying may assist in creating feelings of social connectedness, yet it is still unclear whether that social connection and help happens during the crying experience itself or, perhaps, because crying encourages a person to seek help afterward. Although study participants express willingness to help when shown images of people crying more than when shown images of the same people without tears

(Hendriks et al., 2008; Provine et al., 2009), overall, there has been little research into whether help is actually received by criers. Specifically, it is unclear whether this help is actually functional in improving criers’ emotional state, and furthermore, whether help is received similarly in different cultural contexts.

Crying is more likely to occur in the presence of family and friends than strangers

(Nelson, 2005; Vingerhoets et al., 2000, 1997). Yet it is unclear whether crying among familiar others is more beneficial in terms of help provided and improved emotional state following crying. Of course it is also possible that someone may provide help to a crying CHAPTER 5 122 person, but that the crier perceives it to be unhelpful. Thus far, research is scant on whether social crying is beneficial from the perspective of the crier. However, Bylsma et al. (2011) investigated mood change and its relationship to the number of people present. They found that having another person present was associated with an improved mood, whereas crying alone or with more than one person was associated with worse emotion states following crying. A secondary analysis of the same data set, but instead measuring cathartic crying (i.e., a feeling of emotional release following crying), found that feelings of catharsis were related to socially supportive responses (Bylsma et al., 2008). However, this research did not distinguish between help provided from strangers or persons known to a crier and whether there are cultural differences in how social support is interpreted by the crier.

Therefore, a secondary aim of this research was to understand if the presence of others when a person is crying influences how helped or supported they feel when they cry and whether this influences how emotion is evaluated following crying. We hypothesised that: (H4) similar to previous research findings, the crier’s emotional state would improve most when crying when one person is reported as present; (H5) When the person is known to the crier, criers would be more likely to report that help was provided; (H6) When a crier reports that they have been helped, they would be more likely to report feeling better following crying compared to when no help is reported; (H7) Those who believe that crying in social contexts is unhelpful would be less likely to show an improved emotional state when crying in the presence of others compared to when alone. Specifically, we predicted that those who endorse more beliefs that crying is unhelpful in social contexts would rate a worse emotional state following crying when one or more person is present during crying compared to those crying alone. Given the lack of research across cultures in this area, we also explored possible cross-country differences within these hypotheses. CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 123

All hypotheses and analyses were pre-registered before data collection and can be found on the Open Science Framework (OSF) at https://osf.io/y37xz/. All materials, data files, and analyses can be found on the OSF at https://osf.io/xvdkz/.

Method

Exclusion Criteria

Participants were included in the study if they were aged between 18-40 years, stated that they resided in one of the target countries (see below), and could remember a previous crying experience that was caused by a negative emotional event, which was not caused by something they read in a book or watched on television or in a movie. These criteria were chosen to ensure that participants were comparable across countries in terms of age and crying experiences (i.e., due to a personal negative event). Eighty-three participants did not meet these criteria. Specifically, they indicated that they resided outside of one of the specified countries, their age was outside 18-40 years, they described their last crying experience as caused by watching a film or video clip rather than a personal experience, or stated they could not remember their last crying event at all.

Participants

Following exclusion, a total of 893 participants completed the survey, comprising 508 women and 379 men, and a further five people, four who identified their gender as ‘other’ with some specifying as non-binary or gender queer, and one person who did not answer this question (Mage = 22.81 years, SDage = 5.28). Given the small number of participants who identified their gender as ‘other’, we did not include these participants among analyses that involved gender. However, they were included in all other analyses.

Participants were recruited in Australia, Croatia, the Netherlands, Thailand, and the

United Kingdom. These countries were chosen because they were the home countries of the authors, who shared a mutual interest in crying research. Table 5.1 provides an overview of CHAPTER 5 124 participant characteristics from each country. All participants were screened via their place of residence and aged between 18-40 years and for all countries the most common response for the question asking how long they have lived in that country was that they were born and raised there. Ethics approval for data collection across all countries was granted via the

University of Queensland and the National Statement on Ethical Conduct in Human

Research, and the University of Amsterdam ethics committees. Participants were collected via student samples at each collaborator’s university, collected from the broader community based on their place of residence, or from an online sample (Prolific; www.prolific.ac).

Participants were compensated through course credit at each respective university, and online volunteers were compensated £10 per hour for their participation.

Measures

Triggers. To ensure participants met the criteria for the study, they were asked if the last time they cried was primarily out of sadness, frustration/powerlessness, anger, or something else with an option for free text entry. Participants were also asked the reason they were crying by selecting if they were crying because of something that happened to you, because of something that happened to someone else (e.g., seeing someone hurt in real life),

Table 5.1: Participant demographics for each country, including number of participants by gender, mean (SD) age in years, and the percentage of participants who indicated they were born and raised in that country Country Men Women Non-binary/ Age % born and raised not specified in that country Australia 78 94 3 22.83 (4.39) 56% Croatia 75 79 1 23.59 (6.36) 91% Netherlands 75 150 0 21.05 (4.02) 90.2% Thailand 75 88 2 24.72 (4.88) 94.5% United Kingdom 76 97 1 22.53 (6.07) 80.5%

Note: For participants who indicated they resided in Australia, the second most common response for the length of time lived in Australia after ‘born and raised here’ was 11-15 years. CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 125 or ‘other’ with an option for free text entry. Participants who stated that their reason for crying was due to a trigger that was not caused by something either personal to them (e.g., watching a movie), or because of a positive experience were excluded from analysis.

Emotion Change. Emotion change was measured on a single item asking how participants felt immediately following crying compared to before on a 7-point scale (1= much better, 7= much worse).

Crying Intensity. Intensity of crying was measured using a standardised composite score of two items measuring the length of crying time and amount of tearing during the last crying episode. These items were “how long did you cry for?” measured from 1 (less than 5 minutes) to 5 (continuous [starting and stopping] over a long period of time) and “how intensely did you cry?” rated from 1 (tears in eyes) to 5 (tears down face with vocal sobbing and body movements). These two items were significantly correlated (rs= .39, p < .001) and combined using standardised unit weights.

Crying Frequency and Time Passed. Participants were asked to indicate how long it had been since they last cried. This was measured categorically on a 6-item scale (1 = In the last day, 3 = In the last month, 6 = More than 6 months ago).Participants were also asked to estimate the number of times they cried over the last month for any reason including crying for positive or negative reasons. Importantly, where participants provided a range of crying frequency (e.g., 10-15 times) an average of their estimate was imputed, or the minimum crying experience identified, for example “more than 10 times”, was imputed as ‘11’ times in the last month. This occurred for 65 participants that gave a range or estimate of their crying frequency. If no clear frequency was identified in a response, such as ‘a lot’, then the field was left empty.

Social Context. Participants were asked how many people they were with measured on a 5-item scale (1 = alone, 2 = 1 other person, 3 = 2 other people, 4 = 3 other people, 5 = 4 CHAPTER 5 126 or more other people). To measure if support was received from a person they were with, participants were asked “If you were with one or more people, did you receive support (e.g., emotional, informational, or active help) from them?”. Support was measured on a 4-item scale (1 = yes, 2 = unsure, 3 = no, 4 = not applicable). To understand if the person they were with was known to them, participants were asked “If you were with one or more people, did you know them or were they a stranger?”. Responses were recorded across four-items, 1 = stranger, 2 = acquaintance, 3 = both stranger(s) and acquaintance (s), or 4 = close friend or family member.

Beliefs About Crying Scale (BACS). The BACS consists of three subscales, comprising a total of 14 items (Sharman et al., 2018). Each item is measured on a 5-point scale (1 = Never, 5 = Always), with three subscales, Helpful (7 items; α = .73; e.g., “after crying I feel an emotional release”), Unhelpful-Individual (3 items; α = .65; e.g., “crying makes me feel worse when I’m alone”), and Unhelpful-Social (4 items; α = .78; e.g., “it’s embarrassing when I cry around friends or family”). Given the large sample, small number of items, and lack of hypotheses for Unhelpful-Individual scale, we considered the lower reliability score acceptable for the purposes of this research. Higher scores on the Helpful subscale indicate greater belief in crying as helpful for wellbeing and emotional recovery.

Higher scores on the Unhelpful-Individual scale indicate more beliefs that crying leads to feeling worse when alone. Higher scores on the Unhelpful-Social scale indicate more beliefs that crying is unhelpful in social settings because it can lead to feelings of shame and embarrassment.

Gender roles. Self-ascribed masculinity-femininity was measured using the

Traditional Masculinity Femininity Scale (TMF; Kachel, Steffens, & Niedlich, 2016). This scale measures gender role identity (facets of self-subscribed masculinity-femininity) with items that do not depend on culture and time, as they rely on a person’s own definition of CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 127 masculine/feminine that would be shaped by their culture and experiences. Each of the six items (e.g., “I consider myself as…” and “traditionally, my behaviour would be considered as…”) were measured on a 7-point scale ranging from 1 = very masculine to 7 = very feminine (α = .94).

A measure of gender attitudes (Gender role atitudes; GRA) was also included taken from the wave 6 World Values Survey 2010-2014 (Inglehart et al., 2014). This is a 5-item measure of attitudes toward traditional gender roles including questions such as “A university education is more important for a boy than for a girl” (α = .68). Although all five items were measured, one item was removed to improve overall reliability (“being a housewife is just as fulfilling as working for pay”; ⍺ = .74). Answers were measured on a 4-point scale from 1 = strongly agree to 4 = strongly disagree.

Procedure

Each survey for respective countries was translated and back-translated from English by at least two bilingual speakers of English and Croatian, Dutch, or Thai. Participants were asked to remember their last crying experience that was caused by a negative emotion (e.g., sadness, anger, frustration) that was not caused by something that they read in a book or watched online, on television, or in a movie and to report when they last cried with a free-text entry to describe it if they wished. Regarding this crying experience, participants were asked what caused their crying, how many people they were with and if they received help (i.e., emotional, informational, instrumental), and the intensity with which they cried. Participants were then asked how they felt immediately following crying compared to before and their frequency of crying over the last month. They then completed each of the BACS and gender role scales presented in random order and finally answered demographic questions, relating to their age, country of residence, and gender. Participants self-selected to complete the survey based on the memory of their last crying experience and after being screened for CHAPTER 5 128 relevant age, participants were recruited in their respective countries or online after confirmation of their country of residence. All participants were provided a link to complete this survey online in their own time.

Results

All analyses were conducted in R version 3.5.2 (R Core Team, 2017). Mediation analyses were tested using lavaan (v0.6-3; Rosseel, 2012). The data were first explored to see variations among crying behaviours and evaluations between countries and gender. Overall, women reported crying more frequently, t(752) = 11.9, p < .001, d = .74, and more intensely than men, t(858) = 6.69, p < .001, d = .45. Although men and women’s crying frequency and women’s crying intensity did not significantly differ between countries (all ps > .05), men’s crying intensity showed differences across countries, F(22, 355) = 1.65, p = .033, η2 = .09, with Croatian men showing the lowest crying intensity compared to the other countries sampled (see Table 5.2). Ratings of change in emotional state following crying showed no difference across men and women; t(882) = 0.89, p = .38, d = .06. However, beliefs about crying were significantly different between men and women across all three subscales with women believing that crying was more helpful, t(769) = 5.24, p < .001, d = .36, more unhelpful in social situations, t(761) = 2.27, p = .02, d =.16, and more unhelpful when alone

(t(884) = 3.10, p = .002, d =.21). These ratings showed no differences across countries (all ps

> .05). See Table 5.2 for descriptive statistics.

Crying Relationships and Mediation

Due to the combination of ordinal and continuous variables, and non-normal distribution of the crying intensity variable, initial bivariate correlations were conducted using spearmans rho with α adjusted to < .005 given multiple comparisons. This analysis found significant relationships between the TMF scale, self-reported gender, crying intensity, and the extent to which crying was believed to be helpful. The GRA scale was also correlated CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 129 with gender, TMF, and crying intensity but not related to the BACSHelpful scale, see Table 5.3.

However, emotion states following crying only displayed a relationship with helpfulnes ratings on the BACS.

These relationships indicate that compared to masculine and more traditional gender roles, more feminine and less traditional gender role endorsement was related to greater crying intensity, and beliefs that crying was more helpful overall. Similarly, women were more likely to endorse more feminine and less traditional gender roles, to believe that crying was helpful, and to cry more intensely than men in the sample. Finally, those who believed that crying was helpful were also more likely to feel as though their emotional state improved following crying.

As no relationship was observed between either measure of gender roles and emotional change, a mediation model was only tested with each gender role measure predicting crying intensity and mediated by beliefs that crying is helpful (BACSHelpful; see

Figure 5.1). To ensure time since crying was controlled for, this variable was dichotomised following examination of the distribution of crying into recent crying (crying in the last week or earlier, N = 335) or non-recent crying (crying more than a week ago, N = 559). Because the crying intensity variable displayed a non-normal distribution, bootstrapping was utilised at 1000 samples. These results revealed significant direct effects of both predictors. However, helpfulness, as rated on the BACS, did not mediate the relationship between either self- ascribed gender roles or attitudes toward gender roles and crying intensity when accounting for the time since crying occurred (ab = 0.008, p = .120, CI [-0.003, 0.023]). Secondary mediation analyses were conducted to check if these effects were consistent across countries.

These analyses also found no mediation across countries. See Table 5.4 for all mediational results.

CHAPTER 5 130

Table 5.2: Mean (SD) crying behaviours and evaluations separated by country and gender Country Gender Frequency Emotion Cry BACS BACS unhelpful BACS change Intensity helpful social unhelpful individual All countries Male 1.49 (2.41) 3.41 (1.29) -0.21(0.76) 3.09 (1.03) 3.29 (1.19) 2.00 (0.87) Female 4.6 (5.12) 3.33 (1.26) 0.16 (0.86) 3.44 (0.94) 3.47 (1.07) 2.18 (0.79) Other 4.64 (7.02) 2.86 (1.21) -0.17(0.56) 4.29 (0.87) 2.82 (1.04) 1.57 (0.60) Australia Male 1.62 3.37 -0.28 2.98 3.33 2.11 Female 5.74 3.43 0.20 3.13 3.41 2.11 Other 7.67 3.67 -0.11 3.57 2.92 1.89 Croatia Male 1.33 3.24 -0.40 3.55 3.13 1.84 Female 3.89 3.49 -0.04 4.21 3.46 2.09 Other 1 4 -0.20 5.29 1.5 2 Netherlands Male 1.01 3.45 -0.06 2.96 3.41 2.04 Female 3.91 3.57 0.10 3.24 3.73 2.39 Other ------Thailand Male 1.56 3.23 -0.27 3.06 3.17 1.97 Female 3.73 2.78 0.11 3.5 3.03 1.84 Other 2.25 2 -0.24 4.5 3.38 1.17 United Male 1.9 3.75 -0.03 2.92 3.42 2.04 Kingdom Female 6.02 3.25 0.41 3.4 3.54 2.29 Other 4 1 -0.20 5 2.75 1 Note: Frequency of crying was estimated over the last 30 days; higher scores on emotion change indicate ‘feeling worse’ following crying; higher scores on cry intensity is related to longer crying time with more intense tearing; higher scores on all beliefs scales indicate greater endorsement of the scale.

Table 5.3: Correlations between self-reported gender, gender roles, beliefs that crying is helpful (BACSHelpful), crying intensity, and emotion change following crying

TMF GRA BACSHelpful Cry intensity Emotion change Gender 0.758** 0.311** 0.166** 0.221** -0.035 TMF - 0.331** 0.214** 0.185** -0.040 GES - 0.017 0.150** 0.040 BACSHelpful - 0.096* -0.443** Cry intensity - 0.002 Emotion change - Note: spearman’s rho * p =.005, ** p <.001; Gender = male (1), female (2); TMF = Traditional masculinity/femininity scale; GRA = Gender role attitudes

CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 131

Mediation exploration. Given that there was no correlation found between

BACSHelpful and the GRA scale, to check if the lack of relationship between these variables influenced the mediation, a second mediation model was analyzed using only the TMF scale.

This analysis confirmed the previous results finding that there was no mediation of crying intensity and gender roles measured through the TMF scale via BACSHelpful (ab = 0.009, p =

.20, CI [-0.005, 0.024]).

To explore the assumption that gender roles are a better predictor of crying behaviours than gender alone, a second mediation model was suggested by a reviewer. This model found partial mediation of the relationship between gender and crying intensity, but only through the GRA scale (ab = 0.047, p = .02 CI [0.011, 0.090]) and not the TMF scale

(ab = -0.017, p = .79, CI [-0.150, 0.115]). These results highlight that gender and attitudes toward gender roles likely both play a role in how intensely people cry, c' = 0.366, p < .001,

CI [0.261, 0.476].

Social Context Effects

Correlational analyses and pairwise t-tests were employed to understand the relationship between emotion following crying and the number of people present when crying, with neither finding significant relationships between variables. Specifically, no correlation was revealed between emotion and the number of people present, r = .02, p = .60. Further, when compared against two, three, or more people present when crying, having just one-person present did not appear to influence emotion, all ps > .05. In fact, further comparisons against people who were alone also found no differences in their ratings of emotion when compared to those with people present, ps > 05. No differences were found among these analyses when explored by country.

People were most likely to cry around someone who was known to them (95%), compared to a stranger (5%), and more likely to report receiving help overall (82%), CHAPTER 5 132

Table 5.4: Results overall and across countries testing if the relationship between gender role endorsement and crying intensity is mediated by beliefs that crying is helpful Estimate ab SE p Lower CI Upper CI % Mediation All countries 0.009 0.007 .188 -0.005 0.024 3.96% Australia 0.012 0.015 .44 -0.013 0.047 4.58% Croatia 0.026 0.026 .303 -0.015 0.085 19.85% Netherlands 0.014 0.013 .293 -0.006 0.047 7.82% Thailand 0.031 0.036 .387 -0.033 0.118 17.42% United Kingdom 0.009 0.007 .188 -0.005 0.024 3.96%

Note: ab = indirect effect; SE = standard error; CI = confidence interval; p = significance

compared to not (18%). A chi-square analysis found a significant relationship between reported help and if the persons in the social context were known, 푥2(1, N = 421) = 8.00, p =

0.005. However, as very few people cried around strangers, an additional analysis examined if the strength of the relationship with those present impacted whether help was perceived to be provided, specifically the presence of ‘acquaintances’ compared to a ‘friend/family member’. These results revealed that when a person cries in the presence of someone they know, how well they know that person has no influence on whether help was perceived to be provided, 푥2(1, N = 391) = 0.11, p = 0.74. A t-test was used to explore whether emotion improved when the person was known to the crier that yielded no significant difference but did show a large effect size t(418) = 1.36, p = .18, d = .56. As only six people reported crying around strangers compared to 414 people who reported crying around people they knew, this large effect size suggests the relationship would likely exist in a more equally distributed sample.

To understand whether emotion improved most when help was perceived to be given, compared to when no help was perceived, a t-test was utilised and found a significant difference between the two groups, t(436) = 2.84, p = .005, d = .35. Further exploration found that for those who were in a social situation, there was no interaction of country and help reported (see Figure 5.2). However, there was a significant main effect of country on reported CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 133

Figure 5.2. Plot of emotional change ratings whether help was provided when crying in a social context compared by country. Emotion change scores range from 1= much better, 7= much worse.

emotional change following crying, F(4,428) = 3.06 p = .02, η2 = .03). Further post-hoc comparisons with Tukey corrections revealed that participants living in Thailand rated a better emotional state overall when crying in a social context, compared to those from the

United Kingdom, t(428) = 3.22, p = .01, with no other significant differences emerging.

Finally, a correlational analysis was used to understand the relationship between emotional change following crying and beliefs that crying is unhelpful in social contexts. Results showed that worse emotional change ratings following crying were significantly related to stronger beliefs that crying is unhelpful in social contexts, r = .16, p < .001. This relationship was found to be significant for both people who were exposed to social contexts, r = .16, p = CHAPTER 5 134

.001, and people who were alone, r = .15, p < .001, with a Fisher r-to-z transformation indicating no significant differences between these groups (z = 0.13, p = .89). When analysing these correlations across countries, however, it was revealed that one country (i.e.,

Thailand) was driving this effect with no other countries showing significant relationships either socially or alone with emotional change and beliefs about crying (rThaiSocial= .44, p <

.001, rThaiAlone= .26, p = .008) with a Fisher r-to-z transformation comparing correlations indicating no differences between the two groups (z =1.28, p = .20).

Discussion

The aim of this research was firstly to understand how gender roles may be involved in crying behaviours and emotional effects, and secondly, to gain a better understanding of how the presence of others may impact on participants’ crying experiences. Although this research utilised an international sample, we did not predict cross-cultural differences in the overall relationships between the variables investigated, and largely found none.

Gender Role Endorsement

Corelations. We hypothesised that relationships would exist between gender, attitudes toward gender roles, self-ascribed gender roles, beliefs that crying is helpful, crying intensity, and emotional state following crying. This hypothesis was partially supported by our results. In particular, women were more likely to endorse more feminine and less traditional gender roles, with, as predicted, more self-ascribed feminine gender roles feminine related to greater crying intensity and beliefs that crying was more helpful overall compared to people who identified more masculine gender roles. This result reveals, similar to what was found by Ross and Mirowski (1984) regarding sex-roles, that identification with and attitudes toward gender roles are related to crying behaviours as well as beliefs in the helpfulness of crying. However, we did not find that emotion improvement was related to either self-ascription of gender roles or attitudes toward traditional gender roles. In fact, the CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 135 only variable that was related to emotion change was with beliefs that crying was helpful.

Specifically, those who believed that crying was helpful were also more likely to feel better following crying. Although this relationship may appear circular, it helps to understand the widely held public perception that crying is helpful (Cornelius, 2001). We consider that beliefs about crying’s utility may initially influence how a person evaluates their crying experience. However, it is more likely that these processes reinforce one another. That is, a person’s initial evaluation that crying is helpful would likely be adjusted given a crying experience that leaves the crier feeling worse. Testing the direction of this relationship may be possible in future longitudinal research using experiencing sampling, which would allow the measurement of a number of crying experiences and whether there are related changes to beliefs about crying.

Mediation. Given a lack of relationship between emotion following crying and gender role endorsement, the second hypothesis was not tested in a mediation model. The third hypothesis, however, was also not supported. That is, when accounting for time passed since crying occurred, beliefs that crying is helpful did not account for the relationship between gender role attitudes and self-ascription and intensity of crying. Similar lack of mediated relationships was found for all countries showing that crying intensity seemed to be accounted for by gender roles. To further understand the relationship between gender roles and crying behaviours, they were explored in a mediation model to understand if crying behaviours might be better explained by a person’s self-perception and attitudes toward gender roles than gender alone (Vingerhoets et al., 2000). These results found that both gender and attitudes toward gender roles meaningfully influenced reports of crying intensity.

Specifically, females were more likely to cry more intensely, and the strength of this relationship was partially mediated by females having less traditional attitudes toward gender roles, which in turn predicted more intense crying. CHAPTER 5 136

Taken alongside the correlational results, these effects suggest that a person’s gender and the extent to which they subscribe to their gender role (attitudes and self-perception toward gender roles) is highly related to their behavioural crying responses. However, how crying is evaluated appears instead to be related to our attributions of the helpfulness of crying. This may explain why we have previously seen such large differences in crying frequencies between genders but not large differences in how men and women seem to evaluate their crying experiences (Becht & Vingerhoets, 2002; Sharman et al., 2018)

Crying and Social Help

We also explored the possible social factors involved in crying and emotion change.

Contrary to our predictions, we found that ratings of emotion following crying were not affected by the number of people present. Indeed, this effect persisted even when compared with those who were alone and did not differ across countries. This result is not consistent with previous cross-cultural research that found that the presence of one person during a crying episode was related to the greatest emotional improvement (Bylsma et al., 2008).

These differences in social responses following crying may have arisen from variation in measurement of emotional improvement, with Byslma et al.’s study measuring change following crying using a 3-point scale, compared to the 7-point scale that was used here.

The results showed some support for our fifth hypothesis, that there is a relationship with whether the crier perceives they received help and whether the crier knows the person who helps them. However, comparatively few people indicated that they cried around strangers or people they did not know. Further analyses found that when crying around friends or family compared to acquaintances, there were no differences in whether help was reported by criers. This suggests that simply knowing someone or a group of people, regardless of how strong that relationship is increases a person’s likelihood to cry around them and that no matter the closeness of the relationship with the crier, known others are CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 137 perceived to provide the same amount of help. Finally, we also observed a large effect of emotional improvement when crying around someone the person knew. Although this result was not statistically significant, the large effect size suggests this relationship likely exists, but that our samples split between these groups were not of sufficient size to detect an effect, with only six people indicating they cried in the presence of a stranger who were perceived to help. These results do, however, confirm that people are more likely to cry in front of people known to them, rather than strangers (Vingerhoets et al., 1997). Furthermore, these findings also support the proposed effects of social crying found in previous research, suggesting that crying encourages others to provide help and succour (Hendriks et al., 2008; Hendriks &

Vingerhoets, 2006; Provine et al., 2009). These results also tentatively add to this perspective, indicating that there are distinctions in how help is perceived when provided from strangers or persons known to a crier in whether that help is functional in improving a person’s emotional state.

Support was also found for H6. More precisely, we found that when help was reported by the crier, they tended to rate a better emotional state than those who did not report receiving help from a person present. There were also differences between countries on their reports of emotional change following crying when crying in social situations generally, with participants in the United Kingdom reporting feeling worse following crying compared to those in Thailand. This may suggest a cultural difference among those in the

United Kingdom where there is, perhaps, more shame attached to crying in front of others.

However, with no other research measuring crying and shame comparing these countries, we can only be tentative in our conclusions here. Overall, it appears that help generally does not alter the course of emotion change following crying in most of the countries studied.

Separately, however, some countries did show differentiation in emotion ratings when help was reported, with Thai participants showing the most positive emotion ratings overall when CHAPTER 5 138 help was reportedly provided. This may mean that there are cross-country differences in how people interpret their emotion states when they feel they have been helped, although we do not know whether this is specifically related to crying situations.

We did not find support for our final hypothesis. The relationship between reported emotional state following crying and the belief that crying is unhelpful in social contexts was the same for those who cried in the presence of others and those who cried on their own. That is, regardless of whether people cried alone or in a social context, people reported worse emotional states following crying when they held beliefs that crying is unhelpful in social contexts. However, this effect appeared to be driven exclusively by Thai participants who tended to rate their emotional state as more negative if they believed crying was unhelpful socially, whether they were crying alone or in a social setting. This effect could be explained by the presence of stronger normative beliefs within Thai society that displaying negative emotions is disrespectful as they can cause distress in others (Cassaniti, 2014). All other countries showed no significant relationships between social beliefs and their emotional state following crying. The results from testing Hypotheses 6 and 7 indicate that, cross-culturally, there are differing social and emotional evaluation characteristics that people are exposed to, which likely affect how they interpret the helping behaviours of others and how that intersects with their evaluations of emotions following crying (Heinrichs et al., 2006;

Mesquita, De Leersnyder, & Albert, 2014).

Limitations

The cross-country sample of this research provides good generalisability for the results, with sample sizes within and between countries that allow for strong inferences about potential country differences. Nonetheless, there were several limitations to this research.

First, this study utilised retrospective self-report, which may have biased participants’ responses. As some participants remembered their last crying experience to be longer than six CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 139 months before, it is possible that participants reported their crying experiences more favourably as time passed (Bylsma et al., 2011; Rottenberg, Bylsma, Wolvin, et al., 2008).

Although time since crying was controlled in some analyses, we are still cautious about the conclusions we can make from this research. Further, we note that questions regarding gender that measured global judgements of traits and attitudes may have resulted in greater confirmation of gender stereotypes (Else-Quest, Higgins, Allison, & Morton, 2012).

Additionally, greater conformity to stereotypes may have been primed by questions regarding crying, a stereotypical female behaviour (Jones & Heesacker, 2012). Perhaps future research would benefit from more accurately measuring gender roles using state-based questions.

Second, the average age of our sample was around 23 years old. Although we included an overall age-bracket between 18-40 years, restrictions with sample collections across countries, and attempts to ensure the data were comparable, meant that our samples were much younger. This makes it difficult to generalise our results across age groups, where there may be different influencing social factors, for example, that impact how older or younger adults interpret their crying experience (Blanchard-Fields, Stein, & Watson, 2004;

Zimmermann & Iwanski, 2014).

Thirdly, it is difficult to derive definite conclusions from our results for those participants who reported that they were ‘alone’ when they cried. As one reviewer noted, there may be a myriad of influencing factors, such as the presence of animals or pets and perhaps memories or imagined social others. Future research would benefit greatly from understanding the potential social role of animals or pets during crying, and even how the imagining of social others may impact emotional evaluations following crying.

Conclusion

Overall, explorations into crying behaviour and emotion change across country and gender found that women cried more frequently and more intensely than men across all CHAPTER 5 140 countries, yet ratings of emotion following crying were not different between men and women. Our results strongly suggest that these commonly replicated results based on emotion and crying frequency may be explained by gender, prescription to gender roles and beliefs about whether crying is helpful. That is, self-reported gender and prescription to gender roles appears to be related to the extent to which crying is engaged in, and the intensity of crying

(i.e., greater endorsement of femininity relates to more crying and greater crying intensity).

Conversely, evaluating whether a crying experience results in feeling better or worse appears instead to be related to individual differences in attributions of the helpfulness of crying, rather than gender.

Beliefs about crying were surprisingly different between genders with women not only showing greater beliefs that crying was generally helpful as was found previously, but also beliefs that crying was more unhelpful individually and unhelpful socially, which have formerly shown no differences (Sharman et al., 2018). Although the effects were small, these findings suggest that women may be more sensitive to both positive and negative emotions when crying in different contexts. We do not believe that these results are contradictory but provide further weight to the importance of context specificity. That is, beliefs that crying is helpful are often quite general, such as ‘crying makes me feel better’, and may be more likely to be influenced by popular opinions (Cornelius, 2001). However, when considering crying in specific contexts such as when alone or in a social space, people may be more likely to draw on specific experiences. Further, as women are more frequent criers, they are more likely to have experienced crying in more varying contexts.

These results also provide some further support for the social function of crying. That is, that crying is a strong social signal that encourages help and succour (Gračanin et al.,

2018). Importantly, these results provide a greater understanding of what that helping process means. Specifically, that a person is more likely both to cry and to feel that they will receive CRYING: BELIEFS, GENDER ROLES, AND SOCIAL PRESENCE 141 help around a person that they know, compared to when in the company of a stranger. In addition, that help does not appear to change depending on the extent to which a person is known, with acquaintances and friends/family members perceived to provide equal amounts of support to a crier. Moreover, and notably, when a crier said that they were helped by either instrumental, informational, or physical means, they tended to report a better emotional state following crying than those who did not receive help, but who were around at least one other person. However, it is also important to note that the presence of others has a complex relationship with how people evaluate their emotion following crying, and that the country in which we live may affect how we interpret our crying experience.

CHAPTER 6 142

CHAPTER 6: Discussion and Conclusion

DISCUSSION AND CONCLUSION 143

The understanding that crying benefits a recovery from sadness has been broadly accepted by the lay-public with beliefs about crying held quite strongly across cultures (Becht

& Vingerhoets, 2002). Researchers, too, have hypothesised that crying should bring emotional relief, stress reduction, and a decrease in sympathetic nervous system activation

(Efran & Spangler, 1979; Gross et al., 1994). However, with previously limited research on crying, both in the volume of published research and the strength of methodology, there are considerable gaps in our understanding of crying’s effects. Consequently, my thesis aimed to provide a more comprehensive understanding of crying among adults, with research exploring different functional aspects of crying according to elements of The Multifunctional

Model of Crying, outlined in Chapter 1 and shown again in Figure 6.1. Therefore, to conclude, this chapter will discuss the findings from my thesis and their relationship with elements of the model.

The Multifunctional Model of Crying

I developed the working Multifunctional Model of Crying to attempt to reconcile the different hypothesised functions of crying as was outlined in Chapter 1. The two broader hypotheses on crying, which detail crying as functioning intra- or inter-personally (see

Vingerhoets et al., 2000 for a review), have, more recently, been explored as a joint strategy used in recovery by Miceli and Castelfranchi (2003), who suggested that people use a combination of both individual and social resources to assist with coping. They explained that crying encourages an engagement in independent methods of self-soothing, regulation of distress, and/or as a way to gain help and support from others nearby. Indeed, responses to crying are not unidimensional and involve an interplay of social, cognitive, behavioural, and physiological experience. Thus, given the broader literature that tend to highlight crying as functioning independently within these two broader contexts, The Multifunctional Model of

Crying attempts to reconcile the various complementing hypotheses on the functionality CHAPTER 6 144

Figure 6.1. Updated Multifunctional Model of Crying

of crying into a single model. This working model incorporates both the individual and social components of crying and how they may function independently or in tandem to affect mood, while also incorporating many of the possible influencing factors of crying that have been previously predicted or researched. Building on previous research on adulty crying, the results from this thesis provide some support for The Multifunctional Model of Crying.

Biological Factors

To better understand the role of biological factors in recovery from crying, the research reported in Chapter 3 investigated theoretical perspectives on crying as aiding physical and emotional recovery, as well a first effort to understand if crying acts as a barrier to future pain. The aim of this chapter was to understand if crying involves a protective self- soothing effect, testing the individual response pathway in the model. This research experimentally explored the physiological self-regulating effects of crying by investigating the notion that crying encourages a return to homeostasis and that it may act as a buffer to combat near future stressors, including physical stressors. Results from this study suggest that DISCUSSION AND CONCLUSION 145 the effects of crying on physiology are few, yet complex. Although there were no signals that crying affected hormonal stress levels, or physical coping, there was some indication that crying may be involved in attempts to maintain stability, or return the body to homeostasis, through breathing and heart rate. That is, those who cried showed no changes in breathing rate. However, those who did not cry, and watched either of the neutral or sad video stimuli, actually showed increases in respiration. Furthermore, heart rate changes were also not found for criers or non-criers, with exploratory analyses showing a reduction in heart rate just before crying occurred. These results suggest that crying may be involved in retaining bodily homeostasis, perhaps through feedback with the sympathetic nervous system. This may be because of an unconscious response to crying, or perhaps it is a cognitive effort during crying to control breathing, which likely affects heart rate, in turn.

Overall, this research indicates that for the model, via the individual responses pathway, and particularly when crying alone, biological factors are involved in how a person reacts to crying and may even be involved in conscious efforts to self-regulate through breathing. That is, crying initiates physiological based self-soothing and regulation. However, it is still unclear the extent to which physiological changes during crying are related to appraisal of emotion states. Perhaps the sustained and controlled breathing that follows after the onset of crying results in a regained sense of control of the body and accompanying emotions may be related to an improvement in longer term mood states. Alternatively, these changes are part of a re-interpretation of the event in the future. Unfortunately, this research did not support short-term emotion changes that were related to physiological self-soothing as emotion ratings were the same for those who were exposed to the stimuli but did not cry or show regulation through physiological measures. CHAPTER 6 146

Individual Differences

Further support for the model was provided through a more systematic understanding of individual difference factors in relation to crying evaluations. Chapter 4 reported the development and psychometric validation of the Beliefs About Crying Scale. The aim of this chapter was to futher understand how people appraise and evaluate their crying experiences as part of ‘(re)appraisal’ within the model. The creation of this scale revealed that three overarching beliefs held regarding crying: (1) beliefs that crying is helpful; (2) beliefs that crying is unhelpful when alone; and (3) beliefs that crying is unhelpful in social situations.

These three factors suggest that crying when alone can be viewed as either helpful or unhelpful; however, crying in social situations seems to only be viewed with varying levels of unhelpfulness. From these results it was clear that there are differing beliefs around the effects of crying, with it possible for multiple beliefs to be held by one person. Although this result highlights the complex nature of crying, it also emphasises that these beliefs are highly related to how people evaluate their crying experiences.

To better understand how these beliefs may be associated with changes to negative emotion states, the study reported in Chapter 5 investigated this relationship further through a large cross-country retrospective online survey on crying experiences. Analyses revealed that stronger beliefs that crying was helpful was associated with feeling better after crying and that stronger beliefs that crying was socially unhelpful was associated with feeling worse following crying generally, whether that person was alone or with others.

For the model, these results indicate that individual factors regarding beliefs about crying do impact how a person appraises their crying experience. Indeed, through both individual and social pathways of the model, evaluations of the crying experience depend upon personal beliefs regarding crying. These may be the extent to which a person believes that crying in social situations is embarassing, whether they believe crying alone compounds DISCUSSION AND CONCLUSION 147 negative feelings such as loneliness, and how much a person is inclined to believe that crying is generally helpful to improve how they feel.

Norms and Social Context

The study in Chapter 5 of this thesis also explored if and how socialised gender norms affect our evaluations about crying and to better understand how perceptions of help from social others are related to mood change. Firstly, this chapter was focused on the influencing factors of ‘cultural and social norms’ within the model and how they affect emotion following crying. Through this large cross-cultural sample, analyses revealed that the extent to which people endorse levels of traditional masculinity or femininity did not appear to be involved in emotion changes following crying; this appeared only to be associated with beliefs about crying as helpful. However, gender and self-ascribed gender roles and attitudes were related to whether crying was engaged in and its intensity. Specifically, greater endorsement of self- ascribed traditional femininity and feminine attitudes were associated with more frequent crying and more intense crying. These results imply that gender norm adherence does not seem to influence evaluations of crying but impacts a person’s expression of crying in the first instance, suggesting that social norms may not be as prominent within the model as was predicted.

Secondly, this chapter aimed to understand how the presence of others affects our evaluations of emotion following crying through the social/environmental responses pathway.

Specifically, if people provide instrumental or emotional help to criers as is predicted by the model and whether this help is related to a criers’ emotion change. This research highlighted that a person is more likely both to cry and feel that they received help when around a person that they know compared to the presence of a stranger. Moreover, when a crier reported that they were helped by one or more people, they were more likely to report feeling better following crying than those who did not receive help. CHAPTER 6 148

Overall, the results from Chapter 5 provide some clarification to the model. Firstly, these results revealed that social norms are unlikely to influence crying evaluations, with crying evaluations associated more strongly with individual differences in crying beliefs.

Given this result, cultural and social norms as an influencing factor in the function of crying has been removed from the model (see Figure 6.1). Secondly, these results found that social actors, when responding to crying, tended to act in ways where a crier felt supported in emotional and instrumental ways that result in positive mood changes. This finding supports the social/environmental responses pathway showing that support provided by social actors leads to emotion change.

The Future of the Model

Although the evidence within this thesis provides some support to sections of this working model on how crying functions to influency emotion change, much of it requires further investigation. Principally, more needs to be done to investigate biological factors and their effects on emotional change. The research presented in Chapter 3 was not the first to explore the physiological responses to crying (e.g., Gross et al., 1994), but was a first attempt at exploring possible stress and physical pain involved in crying and recovery. Indeed, this research found some evidence of an attempt for the body to maintain homeostasis for criers.

However, there was no evidence that biological factors were related to the evaluation of emotions following crying, with no differences found between criers and non-criers (who were exposed to the same stimuli) in their affect ratings directly following crying or as time passed. Thus, whether biological changes that maintain homeostasis assist in changing emotion states in a meaningful way, over that of sadness alone, is yet to be determined.

Future research could explore this by measuring participants affective change over a longer period of time and also through qualitative research examining how people’s physiological changes during and after crying might relate to their appraisal of emotion states. DISCUSSION AND CONCLUSION 149

Secondly, Chapters 4 and 5 provided a greater understanding of some of the individual difference factors that influence the evaluation of crying. However, when linking these individual differences (beliefs about crying) to crying behaviours, retrospective self- report was used. As mood following crying is often reported inconsistently, with people often reporting feeling better following crying the farther away they are in time from the event

(Bylsma et al., 2011; Gross et al., 1994; Rottenberg, Bylsma, & Vingerhoets, 2008), it is unclear whether these beliefs about crying are being used to evaluate mood during a crying experience, or just used following crying as a way to understand the experience. Research that compares these beliefs and evaluations of crying directly after a crying experience would thus further our understanding in how individual beliefs about crying might be associated with immediate evaluation of emotion states. This also raises another important question regarding crying more generally. That is, if crying doesn’t immediately result in emotional improvement when alone, then why do we cry so often in private (Bylsma et al., 2008;

Vingerhoets et al., 1997)? To understand crying more generally and to provide further context to the model, future research on crying needs to be explored with criers in naturalistic and private settings. This would likely only be achieved through the use of experience sampling methods, with sampling taking place multiple times a day over a considerable period of time.

Thirdly, although Chapters 4 and 5 attempted to understand the role of individual differences in beliefs about crying, we know very little about other individual factors and their role in how people appraise their crying experience. Although there seems to be associations with other individual difference factors, such as empathy, personality traits

(namely, neuroticism and extraversion), and alexithymia, these have primarily been researched in relation to frequency of crying, rather than crying evaluations (Rottenberg,

Bylsma, Wolvin, et al., 2008). Therefore, research into other individual difference factors and CHAPTER 6 150 their relationship with crying evaluations, through survey or experience sampling designs, would help to create a more comprehensive understanding of these factors within the model.

Further, research from Chapter 4 provided more clear evidence that criers do feel as though others who are present during a crying experience tend to provide them with support.

Again, however, given that this research used retrospective self-report we do not know whether emotion changes following crying would show the same pattern if measured directly following a social crying experience. Within the crying literature overall there have been very few attempts to experimentally research crying when others are present (Labott et al., 1991).

This is an area of crying research that has been neglected and one which requires significantly more exploration if we are to understand how crying truly functions in interpersonal contexts. Research on this phenomenon could be achieved through before/after measurements of emotion with manipulated confederate helping behaviours during a cry- inducing film; or perhaps by using a more naturalistic recruitment of friend dyads with helping behaviours observed while watching cry-inducing stimuli.

Finally, there is still more research needed to understand the extent to which crying may function. Of course, one of the features of crying research to date, including within this thesis, has involved research on crying caused by negative emotions, particularly that of sadness (or helplessness; Miceli & Castelfranchi, 2003). Very little research has been conducted on that of other positive emotions, such as happiness (see Fitzgerald & Vanman,

2018); However, research on other strong emotions, such as anger, have also been limited. A renewed focus on how crying functions when the emotion related to crying is something other than sadness would provide a broader and layered understanding of crying that may deviate from the model proposed here. DISCUSSION AND CONCLUSION 151

Conclusion

Overall, this thesis has found that individual differences, social norms, biological factors, and the social presence of others all play a role in how people evaluate and understand their crying experiences. Despite Darwin’s (1872) conclusion that tears are a meaningless by-product of muscular contractions, this thesis instead reveals that crying, especially from sadness, appears to be purposeful in its evolutionary development. In particular, this evidence assists in our understanding of how crying functions when alone, as a way to initiate self-soothing through conscious and unconscious physiological mechanisms, or in social contexts as a signal to others, and the process by which we evaluate the helpfulness of crying. Moreover, this research provides support towards a new working

Multifunctional Model of Crying. Future research on crying requires innovative solutions to research a phenomenon where its onset, intensity, and evaluation are each so readily influenced by many internal and external factors related to any one person.

152

References

Bagby, R. M., Taylor, G. J., & Parker, J. D. A. (1994). The twenty-item Toronto Alexithymia

scale-II. Convergent, discriminant, and concurrent validity. Journal of Psychosomatic

Research, 38(1), 33–40. https://doi.org/10.1016/0022-3999(94)90006-X

Balsters, M. J. H., Krahmer, E. J., Swerts, M. G. J., & Vingerhoets, A. J. J. M. (2013).

Emotional tears facilitate the recognition of sadness and the perceived need for social

support. Evolutionary Psychology, 11(1), 148–158.

https://doi.org/10.1177/147470491301100114

Bastian, B., Kuppens, P., Hornsey, M. J., Park, J., Koval, P., & Uchida, Y. (2012). Feeling

bad about being sad: The role of social expectancies in amplifying negative mood.

Emotion, 12(1), 69–80. https://doi.org/10.1037/a0024755

Becht, M. C., & Vingerhoets, A. J. J. M. (2002). Crying and Mood Change: A cross-cultural

study. Cognition & Emotion, 16(1), 87–101.

Bell, S. M., & Ainsworth, M. D. (1972). Infant crying and maternal responsiveness. Child

Development, 43, 1171–1190.

Blanchard-Fields, F., Stein, R., & Watson, T. L. (2004). Age differences in emotion-

regulation strategies in handling everyday problems. Journals of Gerontology - Series B

Psychological Sciences and Social Sciences, 59, 261–269.

https://doi.org/10.1093/geronb/59.6.P261

Breuer, J., & Freud, S. (1955). Studies on Hysteria. In J. Strachey (Ed.), The standard edition

of the complete psychological works of Sigmund Freud (Vol 2, pp. 1–305). London:

Hogarth Press.

Brody, L. (2001). Gender, emotion, and the family. Cambridge: Harvard University Press.

Bronstein, P., Briones, M., Brooks, T., & Cowan, B. (1996). Gender and family factors as

predictors of late adolescent emotional expressiveness and adjustment: A longitudinal 153

study. Sex Roles, 34, 739–765. https://doi.org/10.1007/BF01544314

Buchanan, C. M., Eccles, J. S., & Becker, J. B. (1992). Are adolescents the victims of raging

hormones? Evidence for activational effects of hormones on moods and behavior at

adolescence. Psychological Bulletin, 111(1), 62–107. https://doi.org/10.1037/0033-

2909.111.1.62

Bullinger, M., Naber, D., Pickar, D., Cohen, R. M., Kalin, N. H., Pert, A., & Bunney, W. E.

(1984). Endocrine effects of the cold pressor test: Relationships to subjective pain

appraisal and coping. Psychiatry Research, 12, 227–233.

Bylsma, L. M., Croon, M. A., Vingerhoets, A. J. J. M., & Rottenberg, J. (2011). When and

for whom does crying improve mood? A daily diary study of 1004 crying episodes.

Journal of Research in Personality, 45, 385–392.

https://doi.org/10.1016/j.jrp.2011.04.007

Bylsma, L. M., Gračanin, A., & Vingerhoets, A. J. J. M. (2018). The neurobiology of human

crying. Clinical Autonomic Research, 29(1), 63–73. https://doi.org/10.1007/s10286-018-

0526-y

Bylsma, L. M., Vingerhoets, A. J. J. M., & Rottenberg, J. (2008). When is crying cathartic?

An international study. Journal of Social and Clinical Psychology, 27, 1165–1187.

https://doi.org/10.1521/jscp.2008.27.10.1165

Campbell-Sills, L., Barlow, D. H., Brown, T. A., & Hofmann, S. G. (2006). Acceptability

and suppression of negative emotion in anxiety and mood disorders. Emotion, 6, 587–

595. https://doi.org/10.1037/1528-3542.6.4.587

Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the

multitrait–multimethod matrix. Psychological Bulletin, 56, 81–105.

https://doi.org/10.1037/h0046016

Capps, K. L., Fiori, K. L., Mullin, A. S. J., & Hilsenroth, M. J. (2015). Patient crying in 154

psychotherapy: Who cries and why? Clinical Psychology and Psychotherapy, 22, 208–

220. https://doi.org/10.1002/cpp.1879

Carlson, K., & Hall, J. M. (2011). Exploring the concept of manliness in relation to the

phenomenon of crying: A bourdieusian approach. Journal of Holistic Nursing : Official

Journal of the American Holistic Nurses’ Association, 29, 189–197.

https://doi.org/10.1177/0898010110393354

Cassaniti, J. L. (2014). Moralizing emotion: A breakdown in Thailand. Anthropological

Theory, 14, 280–300. https://doi.org/10.1177/1463499614534551

Chaplin, T. M., & Aldao, A. (2013). Gender differences in emotion expression in children: A

meta-analytic review. Psychological Bulletin, 139, 735–765.

https://doi.org/10.1037/a0030737

Chaplin, T. M., Cole, P. M., & Zahn-Waxler, C. (2005). Parental socialization of emotion

expression: Gender differences and relations to child adjustment. Emotion, 5(1), 80–88.

https://doi.org/10.1037/1528-3542.5.1.80

Clean Cut Video. (2014, July 14). I died today by Duke Roberts [Video file]. Retrieved from

https://www.youtube.com/watch?v=bDpLBPcrGBE&t=6s

Coltheart, M. (1980). Iconic memory and visible persistence. Perception & Psychophysics,

27, 183–228. https://doi.org/10.3758/BF03204258

Compare, A., Zarbo, C., Shonin, E., van Gordon, W., & Marconi, C. (2014). Emotional

regulation and depression: A potential mediator between heart and mind.

Cardiovascular Psychiatry and Neurology, 2014. https://doi.org/10.1155/2014/324374

Comrey, A. L., & Lee, H. B. (1992). A first course in factor analysis (2nd ed.). Hillsdale, NJ,

US: Lawrence Erlbaum Associates, Inc.

Cornelius, R. R. (1997). Toward a new understanding of weeping and catharsis? In A. J. J.

M. Vingerhoets, F. J. van Bussel, & A. J. W. Boelhouwer (Eds.), The (non)expression of 155

emotions in health and diease (pp. 303–321). Tilburg, The Netherlands: Tilburg

University Press.

Cornelius, R. R. (2001). Crying and catharsis. In A. J. J. M. Vingerhoets & R. R. Cornelius

(Eds.), Adult crying. A biopsychosocial approach (pp. 199–212). Hove, UK: Routledge.

Costello, A. B., & Osborne, J. W. (2005). Best practices in Exploratory Factor Analysis :

Four recommendations for getting the most from your analysis, 10(7), 1–9.

Cotter, K. N., Prince, A. N., Christensen, A. P., & Silvia, P. J. (2018). Feeling like crying

when listening to music: Exploring musical and contextual features. Empirical Studies

of the Arts, 37, 119–137. https://doi.org/10.1177/0276237418805692

Darwin, C. (1872). The expression of the emotions in man and animals. London, UK: John

Marry.

Davis, M. H. (1980). A multidimensional approach to individual differences in empathy.

JSAS Catalog of Selected Documents in Psychology.

De Fruyt, F. (1997). Gender and individual differences in adult crying. Personality and

Individual Differences, 22, 937–940. https://doi.org/Doi 10.1016/S0191-8869(96)00264-

4

Delp, M. J., & Sackeim, H. A. (1987). Effects of mood on lacrimal flow: Sex differences and

asymmetry. Psychophysiology, 24, 550–556. https://doi.org/10.1111/j.1469-

8986.1987.tb00336.x

Denckla, C. A., Fiori, K. L., & Vingerhoets, A. J. J. M. (2014). Development of the Crying

Proneness Scale: Associations among crying proneness, empathy, attachment, and age.

Journal of Personality Assessment, 96, 619–631.

https://doi.org/10.1080/00223891.2014.899498

Diener, E., & Diener, M. (1995). Cross-cultural correlates of life satisfaction and self-esteem.

Journal of Personality and Social Psychology, 68, 653–663. 156

https://doi.org/10.1037/0022-3514.68.4.653

Dunbar, R. I. M., Baron, R., Frangou, A., Pearce, E., van Leeuwen, E. J. C., Stow, J., … van

Vugt, M. (2012). Social laughter is correlated with an elevated pain threshold.

Proceedings of the Royal Society B: Biological Sciences.

https://doi.org/10.1098/rspb.2011.1373

Efran, J. S., & Spangler, T. J. (1979). Why grown-ups cry. Motivation and Emotion, 3(1),

63–72.

Eid, M., & Diener, E. (2001). Norms for experiencing emotions in different cultures: Inter-

and intranational differences. Journal of Personality and Social Psychology, 81, 869–

885. https://doi.org/10.1037//0022-3514.81.5.869

Else-Quest, N.M., Higgins, A., Allison, C., & Moreton, L. C., Gender differences in self-

conscious emotional experience: A meta-analysis. Psychological Bulletin, 138, 947–

981. https://doi.org/10.1037/a0027930

Fischer, A. H., & Manstead, A. S. R. (2000). The relation between gender and emotions in

different cultures. In A. H. Fischer (Ed.), Gender and emotion: Social psychological

perspectives (pp. 71–94). Cambridge: University Press.

Fischer, A. H., Rodriguez Mosquera, P. M., van Vianen, A. E. M., & Manstead, A. S. R.

(2004). Gender and culture differences in emotion. Emotion, 4(1), 87–94.

https://doi.org/10.1037/1528-3542.4.1.87

Fischer, A. H., & Lafrance, M. (2015). What drives the smile and the tear: Why women are

more emotionally expressive than men. Emotion Review, 7(1), 22–29.

https://doi.org/10.1177/1754073914544406

Fitzgerald, M., & Vanman, E. J. (2018, April). Smile like you mean it. Paper presented at the

Consortium of European Research on Emotion, Glasgow, Scotland.

Fivush, R., Brotman, M. A., Buckner, J. P., & Goodman, S. H. (2000). Gender differences in 157

parent-child emotion narratives. Sex Roles, 42, 233–253.

https://doi.org/10.1023/A:1007091207068

Follenius, M., Brandenberger, G., Hietter, B., Siméoni, M., & Reinhardt, B. (1982). Diurnal

cortisol peaks and their relationships to meals. Journal of Clinical Endocrinology and

Metabolism, 55, 757–761. https://doi.org/10.1210/jcem-55-4-757

Frey, W. H., Desota-Johnson, D., Hoffman, C., & McCall, J. T. (1981). Effect of stimulus on

the chemical composition of human tears. American Journal of Ophthalmology, 92,

559–567. https://doi.org/10.1016/0002-9394(81)90651-6

Frey, W. H., & Langseth, M. (1985). Crying: The mystery of tears. Minneapolis, MN:

Winston Pr.

Frijda, N. H. (1986). The Emotions. In The Emotions. Cambridge: Cambridge University

Press. Retrieved from

http://www.oxfordscholarship.com/view/10.1093/0199253048.001.0001/acprof-

9780199253043

Frymier, A. B., Klopf, D. W., & Ishii, S. (1990). Japanese and Americans compared on the

affect orientation construct. Psychological Reports.

https://doi.org/10.2466/PR0.66.3.985-986

Geliebter, A., Carnell, S., & Gluck, M. E. (2013). Cortisol and ghrelin concentrations

following a cold pressor stress test in overweight individuals with and without night

eating. International Journal of Obesity, 37, 1104–1108.

https://doi.org/10.1038/ijo.2012.166

Gelstein, S., Yeshurun, Y., Rozenkrantz, L., Shushan, S., Frumin, I., Roth, Y., & Sobel, N.

(2011). Human tears contain a chemosignal. Science, 331(6014), 226–230.

https://doi.org/10.1126/science.1198331

Gračanin, A., Bylsma, L. M., & Vingerhoets, A. J. J. M. (2014). Is crying a self-soothing 158

behavior? Frontiers in Psychology, 5(MAY). https://doi.org/10.3389/fpsyg.2014.00502

Gračanin, A., Bylsma, L. M., & Vingerhoets, A. J. J. M. (2018). Why only humans shed

emotional tears: Evolutionary and cultural perspectives. Human Nature, 29, 104–133.

https://doi.org/10.1007/s12110-018-9312-8

Gračanin, A., van Assen, M. A. L. ., Omrčen, V., Koraj, I., & Vingerhoets, A. J. J. M. (2017).

Chemosignalling effects of human tears revisited: Does exposure to female tears

decrease males’ perception of female sexual attractiveness? Cognition and Emotion,

31(1), 139–150. https://doi.org/10.1080/02699931.2016.1151402

Gračanin, A., Vingerhoets, A. J. J. M., Kardum, I., Zupčić, M., Antek, M., & Šimić, M.

(2015). Why crying does and sometimes does not seem to alleviate mood: A quasi-

experimental study. Motivation and Emotion, 39, 953–960.

https://doi.org/10.1007/s11031-015-9507-9

Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and

dysregulation: Development, factor structure, and initial validation of the difficulties in

emotion regulation scale. Journal of Psychopathology and Behavioral Assessment,

26(1), 41–54. https://doi.org/10.1023/B:JOBA.0000007455.08539.94

Green, J. A., Whitney, P. G., & Potegal, M. (2011). Screaming, yelling, whining, and crying:

Categorical and intensity differences in vocal expressions of anger and sadness in

children’s tantrums. Emotion, 11, 1124–1133. https://doi.org/10.1037/a0024173

Gross, J. J., Fredrickson, B. L., & Levenson, R. W. (1994). The psychophysiology of crying.

Psychophysiology, 31, 460–468. https://doi.org/10.1111/j.1469-8986.1994.tb01049.x

Hasson, O. (2009). Emotional tears as biological signals. Evolutionary Psychology, 7, 363–

370.

Heilbrunn, G. (1955). On weeping. Psychoanalytic Quaterly, 24, 245–255.

Heinrichs, N., Rapee, R. M., Alden, L. A., Bögels, S., Hofmann, S. G., Ja Oh, K., & Sakano, 159

Y. (2006). Cultural differences in perceived social norms and social anxiety. Behaviour

Research and Therapy, 44, 1187–1197. https://doi.org/10.1016/j.brat.2005.09.006

Hendriks, M. C. P., Croon, M. A., & Vingerhoets, A. J. J. M. (2008). Social reactions to adult

crying: The help-soliciting function of tears. The Journal of Social Psychology, 148(1),

22–41. https://doi.org/10.3200/SOCP.148.1.22-42

Hendriks, M. C. P., Rottenberg, J., & Vingerhoets, A. J. J. M. (2007). Can the distress-signal

and arousal-reduction views of crying be reconciled? Evidence from the cardiovascular

system. Emotion, 7, 458–463. https://doi.org/10.1037/1528-3542.7.2.458

Hendriks, M. C. P., & Vingerhoets, A. J. J. M. (2006). Social messages of crying faces: Their

influence on anticipated person perception, emotions and behavioural responses.

Cognition & Emotion, 20, 878–886. https://doi.org/10.1080/02699930500450218

Henn, M., Weinstein, M., & Foard, N. (2006). A Short Introduction to Social Research.

London, UK: Sage.

Hervas, G., & Vazquez, C. (2011). What else do you feel when you feel sad? Emotional

overproduction, neuroticism and rumination. Emotion, 11, 881–895.

https://doi.org/10.1037/a0021770

Hesdorffer, D. C., Vingerhoets, A. J. J. M., & Trimble, M. R. (2017). Social and

psychological consequences of not crying: possible associations with psychopathology

and therapeutic relevance. CNS Spectrums, 23, 414–422.

https://doi.org/10.1017/S1092852917000141

Hess, U., David, S., & Hareli, S. (2016). Emotional restraint is good for men only: The

influence of emotional restraint on perceptions of competence. Emotion, 16, 208–213.

https://doi.org/10.1037/emo0000125

Hill, P., & Martin, R. B. (1997). Empathic weeping, social communcation, and cognitive

dissonance. Journal of Social and Clinical Psychology, 16, 299–322. 160

Hofmann, S. G., Carpenter, J. K., & Curtiss, J. (2016). Interpersonal Emotion Regulation

Questionnaire (IERQ): Scale development and psychometric characteristics. Cognitive

Therapy and Research, 341–356. https://doi.org/10.1007/s10608-016-9756-2

Hooper, D., Coughlan, J., & Mullen, M. (2008). Structural Equation Modelling: Guidelines

for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53–

60. https://doi.org/10.1037/1082-989X.12.1.58

Horsten, M., Becht, M. C., & Vingerhoets, A. J. J. M. (1997). Crying and the menstrual

cycle. Psychosomatic Medicine, 59, 102–103.

Horstmanshoff, M. (2014). Tears in ancient and early modern physiology: Petrus Petitus and

Niels Stensen. In D. Kambaskovic (Ed.), Conjunctions of soul, body, and mind from

Plato to the Enlightenment. Dordrecht: Springer.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure

analysis: Conventional criteria versus new alternatives, 6(1), 1–55.

https://doi.org/10.1080/10705519909540118

IBM. (2016). IBM SPSS Advanced Statistics 24. Ibm, 184. http://doi.

org/10.1080/02331889108802322.

Inglehart, R., Haerpfer, C., Moreno, A., Welzel, C., Kizilova, K., Diez-Medrano, J., …

Puranen, B. (Eds.). (2014). World Values Survey: Round six. World Values Survey

Association. Madrid: JD Systems Institute. Retrieved from

http://www.worldvaluessurvey.org/WVSDocumentationWV6.jsp

Ioannou, S., Morris, P., Terry, S., Baker, M., Gallese, V., & Reddy, V. (2016). Sympathy

crying: Insights from infrared thermal imaging on a female sample. PLoS ONE, 11(10),

e0162749. https://doi.org/10.1371/journal.pone.0162749

Jackson, P. L., Meltzoff, A. N., & Decety, J. (2005). How do we perceive the pain of others?

A window into the neural processes involved in empathy. NeuroImage, 24, 771–779. 161

https://doi.org/10.1016/j.neuroimage.2004.09.006 jamovi project. (2018). jamovi (Version 0.9) [Computer Software]. Retrieved from

https://www.jamovi.org.

Jansz, J. (2000). Masculine identity and restrictive emotionality. In A. H. Fischer (Ed.),

Gender and emotion: Social psychological perspectives (pp. 166–168). Cambridge:

Cambridge University Press.

Jiji Press Ltd. (2018). Stressed? Many Japanese schools and companies are encouraging

people to cry to boost mental health. The Japan Times. Retrieved from

https://www.japantimes.co.jp/news/2018/10/13/national/social-issues/crying-drawing-

attention-japanese-schools-companies-mental-detox/#.W9pPJS1L1Tb

John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy: History, measurement, and

theoretical perspectives. In L. A. Pervin & O. P. John (Eds.), Handbook of personality:

Theory and research (pp. 102–138). New York, NY, US: Gilford Press.

https://doi.org/citeulike-article-id:3488537

Johnstone, R. A. (1998). Conspiratorial whispers and conspicuous displays: Games of signal

detection. Evolution, 52, 1554–1563.

Jones, K. D., & Heesacker, M. (2012). Addressing the situation: Some evidence for the

significance of microcontexts with the gender role conflice construct. Psychology of

Men & Masculinity, 13, 294–307. https://doi.org/10.1037/a0025797

Kachel, S., Steffens, M. C., & Niedlich, C. (2016). Traditional masculinity and femininity:

Validation of a new scale assessing gender roles. Frontiers in Psychology, 7(JUL), 1–

19. https://doi.org/10.3389/fpsyg.2016.00956

Kessler, R. C., Andrews, G., Colpe, L. J., Hiripi, E., Mroczek, D. K., Normand, S. L. T., …

Zaslavsky, A. M. (2002). Short screening scales to monitor population prevalences and

trends in non-specific psychological distress. Psychological Medicine, 32, 959–976. 162

https://doi.org/10.1017/S0033291702006074

Kilner, R., & Johnstone, R. A. (1997). Begging the question: Are offspring solicitation

behaviours signals of need? Trends in Ecology & Evolution, 12(1), 11–15.

https://doi.org/https://doi.org/10.1016/S0169-5347(96)10061-6

Koestler, A. (1964). The act of creation. London: Hutchinson.

Kottler, J. A. (1996). The language of tears. San Francisco, CA, CA: Jossey-Bass.

Kraemer, D. L., & Hastrup, J. L. (1988). Crying in adults: Self-control and autonomic

correlates. Journal of Social and Clinical Psychology, 6(1), 53–68.

https://doi.org/10.1521/jscp.1988.6.1.53

Kring, A. M., Smith, D. A., & Neale, J. M. (1994). Individual differences in dispositional

expressiveness: development and validation of the Emotional Expressivity Scale.

Journal of Personality and Social Psychology, 66, 934–949.

Labbé, E., Schmidt, N., Babin, J., & Pharr, M. (2007). Coping with stress: The effectiveness

of different types of music. Applied Psychophysiology and Biofeedback, 32, 163–168.

https://doi.org/10.1007/s10484-007-9043-9

Labott, S. M., Ahleman, S., Wolever, M. E., & Martin, R. B. (1990). The physiological and

psychological effects of the expression and inhibition of emotion. Behavioral Medicine

(Washington, D.C.), 16, 182–189. https://doi.org/10.1080/08964289.1990.9934608

Labott, S. M., & Martin, R. B. (1987). The stress-moderating effects of weeping and humor.

Journal of Human Stress, 13, 159–164.

https://doi.org/10.1080/0097840X.1987.9936809

Labott, S. M., & Martin, R. B. (1988). Weeping: Evidence for a cognitive theory. Motivation

and Emotion, 12, 205–216. https://doi.org/10.1007/BF00993111

Labott, S. M., Martin, R. B., Eason, P. S., & Berkey, E. Y. (1991). Social reactions to the

expression of emotion. Cognition & Emotion, 5, 397–417. 163

https://doi.org/10.1080/02699939108411050

Labott, S. M., & Teleha, M. K. (1996). Weeping propensity and the effects of laboratory

expression or inhibition. Motivation and Emotion, 20, 273–284.

https://doi.org/10.1007/BF02251890

Lakens, D., Scheel, A. M., & Isager, P. (2018). Equivalence testing for psychological

research: A tutorial. Advances in Methods and Practices in Psychological Science, 1,

259–269. https://doi.org/https://doi.org/10.1177%2F2515245918770963

Levine, L. J., & Edelstein, R. S. (2009). Emotion and memory narrowing: A review and goal-

relevance approach. Cognition and Emotion.

https://doi.org/10.1080/02699930902738863

Lögfren, L. B. (1966). On weeping. International Journal of Psychoanalysis, 47, 375–381.

Lombardo, W. K., Cretser, G. A., Lombardo, B., & Mathis, S. L. (1983). Fer cryin’ out loud -

There is a sex difference. Sex Roles, 9, 987–995. https://doi.org/10.1007/BF00290058

Lutz, T. (1999). Crying: The natural and cultural history of tears. New York, NY: W.W.

Norton & Company, Inc.

MacArthur, H. J., & Shields, S. A. (2015). There’s no crying in baseball, or is there? Male

athletes, tears, and masculinity in North America. Emotion Review, 7(1), 39–46.

https://doi.org/10.1177/1754073914544476

Manninen, S., Tuominen, L., Dunbar, R. I., Karjalainen, T., Hirvonen, J., Arponen, E., …

Nummenmaa, L. (2017). Social laughter triggers endogenous opioid release in humans.

The Journal of Neuroscience, 37, 6125–6131,

https://doi.org/10.1523/JNEUROSCI.0688-16.2017

Marston, A., Hart, J. T., Hileman, C., & Faunce, W. (1984). Toward the laboratory study of

sadness and crying. The American Journal of Psychology, 97(1), 127–131.

Martin, R. A. (2012). Is laughter the best medicine? Humor, laughter, and physical health. 164

Psychological Science, 11, 216–220. https://doi.org/10.1111/1467-8721.00204

Martin, R. B., Guthrie, C. A., & Pitts, C. G. (1993). Emotional crying, depressed mood, and

secretory immunoglobulin A. Behavioral Medicine, 19, 111–114.

https://doi.org/10.1080/08964289.1993.9935179

Mesquita, B., De Leersnyder, J., & Albert, D. (2014). The cultural regulation of emotions. In

J. J. Gross (Ed.), Handbook of emotion regulation (pp. 284–301). New York: Guilford

Press.

Miceli, M., & Castelfranchi, C. (2003). Crying: Discussing its basic reasons and uses. New

Ideas in Psychology, 21, 247–273. https://doi.org/10.1016/j.newideapsych.2003.09.001

Miller-Keane, & O’Toole, M. (2003). Miller-Keane encyclopedia and dictionary of medicine,

nursing, and allied health. Gastroenterology Nursing (7th ed.).

https://doi.org/10.1097/00001610-199306000-00012

Mischkowski, D., Crocker, J., & Way, B. M. (2016). From painkiller to empathy killer:

Acetaminophen ( paracetamol ) reduces empathy for pain. Social Cognitive and

Affective Neuroscience, 11,1345–1353. https://doi.org/10.1093/scan/nsw057

Montagu, A. (1959). Natural selection and the origin and evolution of weeping in man.

Science, 130, 1572–1573.

Muris, P., Roelofs, J., Rassin, E., Franken, I., & Mayer, B. (2005). Mediating effects of

rumination and worry on the links between neuroticism, anxiety and depression.

Personality and Individual Differences, 39, 1105–1111.

https://doi.org/10.1016/j.paid.2005.04.005

Nelson, J. K. (2005). Seeing through tears: Crying and attachment. New York: Brunner-

Routledge. https://doi.org/10.4324/9780203955758

Panksepp, J. (1998). Affective neuroscience: The foundations of human and animal emotions.

Affective Science. New York: Oxford University Press. 165

Patel, V. (1993). Crying behavior and psychiatric disorder in adults: A review.

Comprehensive Psychiatry, 34, 206–211. https://doi.org/10.1016/0010-440X(93)90049-

A

Peng, R. C., Yan, W. R., Zhou, X. L., Zhang, N. L., Lin, W. H., & Zhang, Y. T. (2015).

Time-frequency analysis of heart rate variability during the cold pressor test using a

time-varying autoregressive model. Physiological Measurement, 36, 441–452.

https://doi.org/10.1088/0967-3334/36/3/441

Peter, M., Vingerhoets, A. J. J. M., & van Heck, G. L. (2001). Personality, Gender, and

Crying. European Journal of Personality, 15(1), 19–28.

https://doi.org/https://doi.org/10.1002/per.386

Pixar Animation Studios (Rivera, J). (2009). Up.

Pleck, J. H. (1995). The gender role strain paradigm: An update. In A New Psychology of

Men. https://doi.org/10.17763/haer.41.1.91367v0h80051573

Provine, R. R. (2012). Curious Behavior: Yawning, Laughing, Hiccupping, and Beyond.

Cambridge, MA: Belknap Press (Harvard University Press).

Provine, R. R., Krosnowski, K. A., & Brocato, N. W. (2009). Tearing: Breakthrough in

Human Emotional Signaling. Evolutionary Psychology, 7(1), 52–56.

https://doi.org/10.1177/147470490900700107

R Core Team. (2017). R Development Core Team. R: A Language and Environment for

Statistical Computing. Retrieved from https://www.r-project.org/

Roberts, J. E., Gilboa, E., & Gotlib, I. H. (1998). Ruminative response style and vulnerability

to episodes of dysphoria: Gender‚ neuroticism‚ and episode duration. Cognitive Therapy

and Research, 22, 401–423. https://doi.org/10.1023/A:1018713313894

Robyn Arouty. (2014, July 8). Re: I Died Today, By, Duke Roberts. Retrieved from

http://www.robynarouty.com/blog/i-died-today 166

Romans, S. E., Clarkson, R. F., Einstein, G., Kreindler, D., Laredo, S., Petrovic, M. J., &

Stanley, J. (2017). Crying, oral contraceptive use and the menstrual cycle. Journal of

Affective Disorders, 208, 272–277. https://doi.org/10.1016/j.jad.2016.08.044

Rosenberg, E. L. (1998). Level of analysis and the organization of affect. Review of General

Psychology, 2, 247–270. https://doi.org/10.1037/1089-2680.2.3.247

Ross, C. E., & Mirowsky, J. (1984). Men Who Cry. Social Psychology Quarterly, 47, 138–

146. https://doi.org/10.2307/3033942

Rosseel, Y. (2012). lavaan: An R Package for Structural Equation Modeling. Journal of

Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02

Rottenberg, J., Bylsma, L. M., & Vingerhoets, A. J. J. M. (2008). Is crying beneficial?

Current Directions in Psychological Science, 17, 400–404.

https://doi.org/10.1111/j.1467-8721.2008.00614.x

Rottenberg, J., Bylsma, L. M., Wolvin, V., & Vingerhoets, A. J. J. M. (2008). Tears of

sorrow, tears of joy: An individual differences approach to crying in Dutch females.

Personality and Individual Differences, 45, 367–372.

https://doi.org/10.1016/j.paid.2008.05.006

Rottenberg, J., Gross, J. J., Wilhelm, F. H., Najmi, S., & Gotlib, I. H. (2002). Crying

threshold and intensity in major depressive disorder. Journal of Abnormal Psychology,

111(2), 302–312. https://doi.org/10.1037/0021-843X.111.2.302

Saifon Th Angel. (2014, November 15). The sisters – and the true meaning of hair – Thai ads

[Video file]. Retrieved from https://www.youtube.com/watch?v=pLDUY-j5-kc

Sachs, L. J. (1973). On Crying, Weeping and laughing as defences against sexual drives, with

special consideration of adolescent giggling. International Journal of Psycho-Analysis,

54, 477–484.

Schaffer, C. E. (1993). The role of adult attachment in the experience and regulation of 167

affect. Yale University.

Schwabe, L., & Schächinger, H. (2018). Ten years of research with the Socially Evaluated

Cold Pressor Test: Data from the past and guidelines for the future.

Psychoneuroendocrinology, 92, 155–161.

https://doi.org/10.1016/j.psyneuen.2018.03.010

Sharman, L. S., Dingle, G. A., & Vanman, E. J. (2018). Does crying help? Development of

the beliefs about crying scale (BACS). Cognition and Emotion, 33, 722–736.

https://doi.org/10.1080/02699931.2018.1488243

Sharman, L. S., Vanman, E. J., & Scambler, J. (2015). The Social Function of Tears in

Crying. Paper presented at the 25th Annual Conference of the Australasian Society for

Psychophysiology, Sydney, Australia.

Simons, G., Bruder, M., van der Löwe, I., & Parkinson, B. (2013). Why try (not) to cry:

Intra- and inter-personal motives for crying regulation. Frontiers in Psychology, 3(JAN),

1–9. https://doi.org/10.3389/fpsyg.2012.00597

Srivastava, S., John, O. P., Gosling, S. D., & Potter, J. (2003). Development of personality in

early and middle adulthood: Set like plaster or persistent change? Journal of Personality

and Social Psychology, 84, 1041–1053. https://doi.org/10.1037/0022-3514.84.5.1041

Steiger, J. H. (2007). Understanding the limitations of global fit assessment in structural

equation modeling. Personality and Individual Differences, 42, 893–898.

https://doi.org/10.1016/j.paid.2006.09.017

Stein, M. B., & Stein, D. J. (2008). Social anxiety disorder. Lancet, 371, 1115–1125.

https://doi.org/10.1016/S0140-6736(08)60488-2

Stone, A. A., Schwartz, J. E., Smyth, J., Kirschbaum, C., Cohen, S., Hellhammer, D., &

Grossman, S. (2001). Individual differences in the diurnal cycle of salivary free cortisol:

A replication of flattened cycles for some individuals. Psychoneuroendocrinology, 26, 168

295–306. https://doi.org/10.1016/S0306-4530(00)00057-3

Story Corps (Rauch, M., & Jacobs, L.). (2011). Always a family [Video file]. Available from

https://storycorps.org/animation/always-a-family/

Tan, K., Jarnecke, A. M., & South, S. C. (2017). Impulsivity, communication, and marital

satisfaction in newlywed couples. Personal Relationships, 24, 423–439.

https://doi.org/10.1111/pere.12190

Thoma, M. V., Scholz, U., Ehlert, U., & Nater, U. M. (2012). Listening to music and

physiological and psychological functioning: The mediating role of emotion regulation

and stress reactivity. Psychology & Health, 27, 227–241.

https://doi.org/10.1080/08870446.2011.575225

Thompson, E. R. (2007). Development and validation of an internationally reliable short-

form of the positive and negative affect schedule (PANAS), 38, 227–242.

https://doi.org/10.1177/0022022106297301

Timmers, M., Fischer, A. H., & Manstead, A. S. R. (1998). Gender differences in motives for

regulating emotions. Personality and Social Psychology Bulletin, 24, 974–985.

https://doi.org/10.1177/0146167298249005

Transport Accident Commission (TAC) Victoria. (2008). Pictures of you [Video file].

Available from http://picturesofyou.com.au/commercial.php

Walter, J. G. (2018). The adequacy of measures of gender roles attitudes: A review of current

measures in omnibus surveys. Quality & Quantity, 52, 829–848.

https://doi.org/10.1007/s11135-017-0491-x van Hemert, D. A., van de Vijver, F. J. R., & Vingerhoets, A. J. J. M. (2011). Culture and

crying: Prevalences and gender differences. Cross-Cultural Research, 45, 399–431.

https://doi.org/10.1177/1069397111404519 van Tilburg, M. A. L., Becht, M. C., & Vingerhoets, A. J. J. M. (2003). Self-reported crying 169

during the menstrual cycle: Sign of discomfort and emotional turmoil or erroneous

beliefs? Journal of Psychosomatic Obstetrics and Gynecology, 24, 247–255.

https://doi.org/10.3109/01674820309074689 van Tilburg, M. A. L., Unterberg, M. L., & Vingerhoets, A. J. J. M. (2002). Crying during

adolescence: the role of gender, menarche, and empathy. British Journal of

Developmental Psychology, 20, 77–87. https://doi.org/10.1348/026151002166334

Vanman, E. J., Sharman, L. S., & Chantra, M. (Under Review). How his and her tears are

perceived differently: Gender, context, and the Tear Effect.

Vega-Redondo, F., & Hasson, O. (1993). A game-theoretic model of predator-prey signaling.

Journal of Theoretical Biology. https://doi.org/10.1006/jtbi.1993.1089

Verissimo, R., Mota-Cardoso, R., & Taylor, G. J. (1998). Relationships between alexithymia,

emotional control, and quality of life in patients with inflammatory bowel disease.

Psychotherapy and Psychosomatics, 67, 75–80.

https://doi.org/10.1159/10.1159/000012263

Vingerhoets, A. J. J. M. (2013). Why only humans weep: Unravelling the mysteries of tears.

Oxford: Oxford University Press.

Vingerhoets, A. J. J. M., & Becht, M. C. (1996). The ISAC Study: Some preliminary results.

In the International Study of Adult Crying (ISAC) Symposium. Tilburg, The Netherlands,

The Netherlands.

Vingerhoets, A. J. J. M., Boelhouwe, A. J. W., van Tilburg, M. A. L., & van Heck, G. L.

(2001). The situational and emotional context of adult crying. In A. J. J. M. Vingerhoets

& R. R. Cornelius (Eds.), Adult crying. A biopsychosocial approach (pp. 71–90). Hove

(UK): Brunner-Routledge.

Vingerhoets, A. J. J. M., & Bylsma, L. M. (2015). The riddle of human emotional crying: A

challenge for emotion researchers. Emotion Review. 170

https://doi.org/10.1177/1754073915586226

Vingerhoets, A. J. J. M., Bylsma, L. M., & Rottenberg, J. (2009). Crying: A biopsychosocial

phenomenon. In T. Fogen (Ed.), Tears in the Graeco-Roman World (pp. 439–475).

Berlin & New York: de Gruyter.

Vingerhoets, A. J. J. M., Cornelius, R. R., van Heck, G. L., & Becht, M. C. (2000). Adult

crying: A model and review of the literature. Review of General Psychology, 4, 354–

377. https://doi.org/10.1037/1089-2680.4.4.354

Vingerhoets, A. J. J. M., & Kirschbaum, C. (1997). Crying, mood, and cortisol.

Psychosomatic Medicine, 59, 92–93.

Vingerhoets, A. J. J. M., & Scheirs, J. (2000). Sex differences in crying: Empirical findings

and possible explanations. In A. H. Fischer (Ed.), Gender and emotion: Social

psychological perspectives (pp. 143–166). Cambridge, UK: Cambridge: University

Press.

Vingerhoets, A. J. J. M., van de Ven, N., & van der Velden, Y. (2016). The social impact of

emotional tears. Motivation and Emotion, 40, 455–463. https://doi.org/10.1007/s11031-

016-9543-0

Vingerhoets, A. J. J. M., van den Berg, M. P., Kortekaas, R. T. J., van Heck, G. L., & Croon,

M. A. (1993). Weeping: Associations with personality, coping, and subjective health

status. Personality and Individual Differences, 14(1), 185–190.

https://doi.org/10.1016/0191-8869(93)90188-9

Vingerhoets, A. J. J. M., van Geleuken, A. J. M. L., van Tilburg, M. A. L., & van Heck, G. L.

(1997). The Psychological Context of Crying: Towards a Model of Adult Crying. In A.

J. J. M. Vingerhoets, F. van Bussel, & A. Boelhouwer (Eds.), The (non)expression of

Emotions in Health and Disease (pp. 323–336). Tilburg: Tilburg University Press.

Retrieved from http://arno.uvt.nl/show.cgi?fid=48635 171

Wheeler, K., Greiner, P., & Boulton, M. (2005). Exploring alexithymia, depression, and

binge eating in self-reported eating disorders in women, 41, 114–123.

Williams, D. G. (1982). Weeping by adults: Personality correlates and sex differences. The

Journal of Psychology, 110, 217–226. https://doi.org/10.1080/00223980.1982.9915343

Zahn-Waxler, C., Friedman, R. J., & Cole, P. M. (1996). Japanese and United States

preschool children’s responses to conflict and distress. Child Development, 67, 2462–

2477. https://doi.org/10.1111/j.1467-8624.1996.tb01868.x

Zeifman, D. M. (2001). An ethological analysis of human infant crying: Answering

Tinbergen’s four questions. Developmental Psychobiology, 39, 265–285.

https://doi.org/http://dx.doi.org/10.1002/dev.1005

Zeifman, D. M., & Brown, S. A. (2011). Age-related changes in the signal value of tears.

Evolutionary Psychology, 9, 313–324.

Zeman, J. L., & Garber, J. (1996). Display rules for anger, sadness, and pain. Child

Development, 67, 957–973.

Zimmermann, P., & Iwanski, A. (2014). Emotion regulation from early adolescence to

emerging adulthood and middle adulthood: Age differences, gender differences, and

emotion-specific developmental variations. International Journal of Behavioral

Development, 38, 182–194. https://doi.org/10.1177/0165025413515405

172

Appendices

Appendix A

Supplementary Materials for Chapter 3

Crying Frequency Graphs

The frequency graphs assessed for a subset (half) of 34 crying participants across each of the individual videos at 10-second epochs. Participants were either marked as ‘crying’ or ‘not crying’ at each segment. Potential carry-over crying for the initial 20 seconds of each video, excluding the first, was agreed by researchers not to be included.

Each of the videos are listed below and are can be found compiled within the materials at this link: https://osf.io/n3u95/

Video 1 – Dogs last day (Clean Cut Video, 2014; video created from blog post by Robyn Arouty, 2014) Video 2 – Sister (Saifon Th Angel, 2014) Video 3 – Up opening scene (Rivera, 2009) Video 4 – Always a family (Rauch & Jacobs, 2009) Video 5 – Pictures of you (TAC, 2008)

173

174

175

Supplementary Exploratory Analyses for Chapter 3

Measures

Crying proneness scale (CPS). The crying proneness scale (Denckla et al., 2014) is a measure of an individual’s disposition to crying using examples of positive and negative antecedents. The CPS consists of 27 items measured on a 7-point scale (1= very unlikely, 7= very likely). The CPS has an overall rating of crying proneness and is also separated into four subscales of (1) attachment tears (e.g., “the birth of a child”), (2) societal tears (e.g., “a conflict in a group”), (3) sentimental tears (e.g., “when someone is taken in by a group”), and

(4) compassionate tears (e.g., “when someone is seriously hurt”). Reliability for each scale was calculated and found to range from good to excellent (all α > .77)

Interpersonal reactivity index (IRI). Participants were asked to complete the IRI

(Davis, 1980), a measure of trait empathy, consisting of four scales: (1) perspective taking

(α= .81), (2) fantasy (α= .76), (3) empathic concern (α= .80), and (4) personal distress

(α= .76). Overall, the IRI consists of 28-items measured on a 5-point scale (1 = does not describe me well, 5 = describes me very well). The perspective taking scale includes items relating to the ability to adopt others points of view (e.g., “I try to look at everybody’s side of a disagreement before I make a decision”); the fantasy scale measures the ability to imaginatively transpose oneself into the feelings and actions of fictitious characters in books, movies, and plays (e.g., “I really get involved with the feelings of the characters in a novel”); the empathic concern subscale looks at feelings of concern for others in distress (e.g., “I am quite often touched by things that I see happen”); and the personal distress scale measures personal unease in social settings (e.g., I tend to lose control during emergencies”).

Interpersonal Emotion Regulation Questionnaire (IERQ). The IERQ (Hofmann et al., 2016) consists of 20 items with five items in each of four sub-scales: (1) enhancing positive affect (e.g., “when I feel elated, I seek out other people to make them happy”), (2) 176 perspective taking (e.g., “Having people remind me that others are worse off helps me when

I’m upset”), (3) soothing (e.g., “When I feel sad, I seek out others for consolation”), and (4) social modelling (e.g., “It makes me feel better to learn how others dealt with their emotions”). Overall, the scale shows strong psychometric characteristics with reliability for all scales between .81 and .91 in our sample.

Emotion Regulation Questionnaire (ERQ). The ERQ has two subscales representing emotion regulation strategies of cognitive reappraisal and expressive suppression. The cognitive reappraisal facet contains items focusing on changing emotions via reappraisal (e.g., “I control my emotions by changing the way I think about the situation

I’m in”). The expressive suppression items focused on the suppression of emotional behaviours (e.g., “When I am feeling negative emotions I am careful not to express them”).

Our sample found good reliability for the cognitive reappraisal (α = .77) facet. However, the expressive suppression scale showed lower reliability (α > .54), so results from this subscale were not analysed.

Social Expectancies Scale (SES). The SES measures how acceptable an individual feels it is to express individual emotions (Bastian et al., 2012). For this study, 13 items were modified for crying and rated on a 9-point ranging from 1 (strongly disagree) to 9 (strongly agree). These items asked participants to rate their self-evaluation of crying (two items; e.g.,

“Crying makes me dislike myself”; α = .76), their personal expectations of crying (five items; e.g., “Crying is normal”; α = .82), and perceived social expectancies (six items; e.g., “It’s important that others don’t see me as a crier”; α = .71), with good reliability in our sample. 177

Results

Self-report measures and experimentally induced crying

Below, we present between-group comparisons on established measures of crying proneness and emotion regulation. See Table A1 for means, standard deviations, and correlations with emotional responses to the stimuli.

Group comparisons. For the Crying Proneness Scale (CPS) a significant difference was found between Groups for ratings on the total CPS, F(2, 194) = 6.95, p = .001, η2 = .07, and subscales for attachment, F(2, 194) = 5.17, p = .007, η2 = .06, sentiment, F(2, 194) = 4.88, p = .009, η2 = .05, and compassion F(2, 194) = 9.23, p < .001, η2 = .09. Follow-up comparisons found differences between the Neutral and No-Cry, and Cry and No-Cry, groups for the each of the subscales (all ps < .05).

The IRI displayed differences for the fantasy scale F(2, 194) = 6.66, p = .002, η2 = .06, and the empathic concern subscales, F(2, 194) = 5.38, p = .005, η2 = .05. Follow-up comparisons revealed a difference only between the Cry and No-cry groups for both subscales (both ps < .01).

The IERQ only revealed differences between groups for the enhancing positive affect subscale, F(2, 194) = 6.29, p = .002, η2 = .07. Follow-up comparisons displayed differences between the Cry group and both the Neutral, p = .03, and the No-cry group, p = .003, but no differences were revealed for the Neutral vs. No-Cry groups, p = 1.00. Finally, Neither the

ERQ subscale for reappraisal, F(2, 193) = 0.14, p = .87, η2 = .001, or any subscale from the expectancies scale displayed any differences between groups (self-evaluation: F(2, 194) =

1.34, p = .27, η2 = .01; personal expectancies: F(2, 194) = 0.64, p = .53, η2 = .007; social expectancies: F(2, 194) = 0.29, p = .75, η2 = .003).

Correlations. To explore whether these established measures are correlated with crying, correlations were explored for participants in the Cry or No-cry groups (N = 128). 178

These were explored for cry-ratings and overall change in cortisol (AUCI) compared to each self-report measure and participant age. Relationships were found between self-ratings of emotional response and the CPS for total, attachment, sentimental, and compassionate crying, with higher scores on each relating to increased reports of tearing. This same relationship was found for the IERQ, with those who rated more emotional responses during video watching indicating greater use of others to enhance positive affect. Again, higher scores on the IRI for the fantasy scale, perspective taking, and empathic concern were also related to increased emotional responses during video watching. No overall relationships were found between cortisol levels and self-report measures except for a correlation with empathic concern on the

IRI, with higher scores on empathic concern related to reductions in cortisol levels. No relationship was found with participant age. See Table A1 for correlations.

To explore these results further, a standard multiple linear regression was employed to understand the proportion of emotional responses explained by the self-report data where relationships were previously found among the CPS, the IERQ, and the IRI (see Table A2).

Each of the predictor variables was found to have significant zero-order correlations with the emotional response to the film stimuli, but only the CPS total scale (p = .04), the IERQ subscale for enhancing positive affect (p = .005), and the fantasy scale for the IRI (p = .005) were significant predictors of emotional response in the regression model. This five-predictor model was able to account for 25% of the variance in emotional responses to the video stimuli for those who had watched the cry-eliciting stimuli, F(5, 123) = 8.10, p < .001, η2

= .25.

Brief Discussion

Correlational results from this study suggest that people who actually cried were more likely to report that they cry easily and were also more likely to use others to enhance already felt positive emotions. The criers also showed higher scores on empathic measures for fantasy, 179 perspective taking, and empathic concern, implying that the videos used to elicit cry- emotional responses were linked to participants’ ability to empathise with the characters in the emotional video stimuli. Furthermore, these variables accounted for 25% of the variance in emotional responding across those who were exposed to the sad stimuli.

Notably, however, the only correlate of cortisol reduction, was not emotional response, but rather, greater levels of empathic concern. 180

Table A1

Means (SDs) of cortisol and self-report measures across groups and exploratory correlations with sad-criers and sad-non criers reported intensity of emotional responses to the video stimuli

Means (SDs) Correlations Cry No-cry Neutral Emotional Cortisol

Response AUCI

Cortisol AUCI -32.54 (62.32) -28.26 -27.53 -.044 (54.38) (86.43) - Age .011 .078 CPS Total 90.61 (24.90) 74.85 (23.46) 88.29 (28.57) .370** -0.093 Attachment 25.52 (7.26) 21.87 (6.62) 25.35 (7.70) .348** -0.071 Societal 7.11 (3.18) 6.57 (3.22) 7.38 (3.92) .127 -.039 Sentimental 23.14 (9.55) 18.52 (7.90) 22.55 (9.56) .301** -.107 Compassionate 34.83 (8.85) 27.88 (9.03) 33.00 (10.56) .393** -.079 IERQ Enhancing Positive Affect 21.67 (3.02) 19.66 (3.86) 20.14 (3.40) .339** -.103 Perspective Taking 12.42 (4.43) 12.05 (4.53) 11.46 (4.37) .007 -.123 Soothing 14.79 (5.05) 14.18 (5.33) 15.18 (6.08) .146 -.070 Social Modelling 16.35 (4.38) 16.12 (4.40) 16.25 (4.68) .045 -.057 ERQ Reappraisal 30.14 (6.18) 30.26 (5.56) 30.66 (6.24) -.017 -.015 IRI Fantasy Scale 21.27 (4.55) 18.16 (4.79) 19.66 (5.30) .349** -.095 Empathic Concern 21.73 (4.68) 19.25 (3.57) 20.62 (4.62) .285** -.247* Perspective Taking 19.28 (4.64) 17.64 (4.06) 18.55 (5.45) .191* -.007 Personal Distress 13.20 (5.28) 13.72 (4.08) 14.20 (5.06) .001 -.107 Expectancies Scale 3.92 (1.17) 4.07 (1.14) 3.93 (1.14) Self-Evaluation 2.89 (1.86) 3.25 (1.89) 2.72 (1.80) -.004 -.088 Personal Expectation 2.82 (1.57) 3.14 (1.61) 2.90 (1.81) -.138 .094 Social Expectancies 4.83 (1.48) 4.64 (1.38) 4.75 (1.45) .068 .072 Note: *p < .05; **p ≦ .001; Emotional responses were rated on a 7-point scale ranging from 1 = ‘no emotion’, 3

= ‘moved’, 7= ‘sobbing’; AUCI = area under the curve with respect to increase; CPS = Crying Proneness Scale; IERQ = Interpersonal Emotion Regulation Questionnaire; ERQ = Emotion Regulation Questionnaire; IRI = Interpersonal Reactivity Index. 181

Table A2 Self-report data related to emotional reactivity to video stimuli (N = 128).

Zero-order r β sr2 b t Variable CPS EPA FS EC PT ER CPS - .365** .192 .02 .016 2.07* EPA .316** - .339** .247 .05 .143 2.85* FS .378** .116 - .350** .254 .06 .107 2.84* EC .472** .393** .428** - .290** -.036 .001 -.017 -0.36 PT .207* .200* .134 .366** - .191* .081 .008 .038 0.96 Mean 83.32 20.72 19.77 20.64 18.51 4.67 Intercept = .85 (SD) (25.49) (3.59) (5.06) (4.45) (4.49) (2.51) R2 = 0.25

Note: *p < .05; ** p < .001; CPS = Crying Proneness Scale Total; EPA = Enhancing Positive Affect subscale on the Interpersonal Emotion Regulation Scale; Subscales from the Interpersonal Reactivity Index: FS = Fantasy Scale, EC = Empathic Concern, PT = Perspective Taking; ER = Emotional Response.

182

Appendix B

Open-ended questions used for item generation in Chapter 4

1. Do you find that there’s a difference (positive, negative, or neutral) between how you

feel right before crying compared to just after? Please describe using full sentences

For the following questions try to think about the last time you cried, or when you have cried in one of the situations given. If you have not cried recently, try to imagine yourself crying in the situations given. Please try to answer truthfully and in full sentences by typing your response in the boxes provided.

2. Do you find that there’s a difference (positive, negative, or neutral) between how you

feel right before crying compared to just after?

3. Thinking about a time you recently cried, describe what you believe crying did for

you?

4. What do you believe crying does for you when you are alone?

5. What do you believe crying does for you when you are with others (at least one other

person)?

6. Please think about and describe what you believe crying would do or has done to

help, hinder, or have no effect on you personally in each of the following situations:

a. Personal suffering

b. Physical pain

c. Feelings of separation and/or loss

d. Failure

e. Feelings of anger

f. Feelings of guilt 183

Appendix C

Initial item pool used to generate the BACS

Instructions: The following questions ask about what you believe crying does for you. Please answer on the scale how often each statement is true for you. Scoring: 1= never, 3 = about half the time, 5 = always

1. After crying I feel an emotional release. 2. I feel comfortable crying when I'm alone. 3. I don't feel judged when I cry alone. 4. I feel safe when I cry alone. 5. Crying makes me feel better. 6. I feel less stressed after I have cried. 7. Crying helps me to process my emotions. 8. In the long run, I know that I'll feel better because I have cried. 9. I feel worse immediately after I cry. 10. Crying makes me feel more human. 11. Crying helps me to express how I'm feeling without words. 12. I feel support when I'm crying around friends or family. 13. When I cry because I am in physical pain, it's a signal to others that I need help. 14. When I cry around or with friends or family I feel a sense of connection with them. 15. I feel empowered when I cry in front of other people. 16. Crying helps me to process failure. 17. Crying helps when I'm feeling overwhelmed. 18. Crying helps me to get through feelings of physical pain. 19. Crying distracts me from my pain (physical or mental). 20. Crying makes me feel worse when I'm alone. 21. When I'm by myself, crying makes me feel lonely. 22. Crying makes me feel immediately better. 23. In the long run, I will feel worse because I have cried. 24. Crying around others makes me feel me vulnerable. 25. I feel ashamed when I cry around people who are not my friends or family members. 26. I feel judged when I cry around co-workers. 27. It’s embarrassing when I cry around friends or family. 28. I feel like I am less of a person when I cry. 184

29. Crying makes my feelings of suffering worse. 30. When I feel angry, crying makes me feel angrier. 31. When I'm angry, crying makes me feel like I am giving in to the situation. 32. Crying because I have failed at something makes me feel worse. 33. If I have done something wrong, crying makes me feel guiltier. 34. Crying helps to get what I want. 35. Crying does nothing for me (neither better or worse). 36. If I cry when others are around, crying doesn’t change how I feel. 37. When people I’m with are crying out of grief, I feel I should cry too. 38. After I have cried, I don’t feel any different. 39. Crying is something I feel I need to 'get out of the way'. 40. If I cry during an argument, it reduces the intensity of the argument.

185

Appendix D

Ethics Approvals Chapter 2

186

Chapter 3

187

Chapter 4

188

Chapter 5