Master

Same-sex attraction in homophobic men: The role of impulsive processes

GUERRA MENESES, Adrianne

Abstract

While the underlying causes of are not fully understood, one theory claims it stems from an unconscious or denied attraction to the same-sex. A study by Adams, Wright, and Lohr (1996) found evidence of this same-sex attraction in homophobic men, but other studies have shown mixed results. Drawing on dual-process models, this study tested the assumption that in homophobic men, sexual interest in homosexual stimuli depends on their specific impulses towards these stimuli. Thirty-eight heterosexual men (Mage = 22 years, SD= 4.74) first completed a questionnaire measuring negative attitudes towards homosexuals. Impulsive tendencies to approach homosexual stimuli (IAH) were then evaluated via a manikin task. Next, participants completed a picture-viewing task with simultaneous eye-tracking recording to assess viewing time of visual areas of interest (i.e., face and body). In our main hypothesis, we expected that in highly homophobic men, IAH would predict viewing time of pictures of homosexual couples, but not of heterosexual couples. Overall, the results of the linear mixed model analyses supported this hypothesis: [...]

Reference

GUERRA MENESES, Adrianne. Same-sex attraction in homophobic men: The role of impulsive processes. Master : Univ. Genève, 2015

Available at: http://archive-ouverte.unige.ch/unige:81099

Disclaimer: layout of this document may differ from the published version.

1 / 1

Same-sex attraction in homophobic men: The role of impulsive processes

MÉMOIRE RÉALISÉ EN VUE DE L’OBTENTION DE LA MAÎTRISE UNIVERSITAIRE EN PSYCHOLOGIE

ORIENTATIONS PSYCHOLOGIE COGNITIVE PSYCHOLOGIE CLINIQUE

PAR

Adrianne Guerra Meneses

DIRECTEUR DU MEMOIRE Julien Chanal Co-directeur Boris Cheval

JURY Boris Cheval Julien Chanal Roland Maurer

GENÈVE, décembre 2015

UNIVERSITÉ DE GENÈVE FACULTÉ DE PSYCHOLOGIE ET DES SCIENCES DE L’ÉDUCATION SECTION PSYCHOLOGIE

Acknowledgements

I would like to express my sincere gratitude to all the people who have made this work possible. I would first like to thank my supervisor, Boris Cheval, for his invaluable guidance and encouragement throughout the entire process. Next, I would like to thank Julien Chanal and Roland Maurer for taking the time to read and evaluate my work. I would also like to thank my research partner, Gwendoline Peyratout, for her 2 years of collaboration. Last but not least, a special thanks to Duarte for his unconditional and endless support.

Abstract

While the underlying causes of homophobia are not fully understood, one theory claims it stems from an unconscious or denied attraction to the same-sex. A study by Adams, Wright, and Lohr (1996) found evidence of this same-sex attraction in homophobic men, but other studies have shown mixed results. Drawing on dual-process models, this study tested the assumption that in homophobic men, sexual interest in homosexual stimuli depends on their specific impulses towards these stimuli. Thirty-eight heterosexual men (Mage = 22 years, SD= 4.74) first completed a questionnaire measuring negative attitudes towards homosexuals. Impulsive tendencies to approach homosexual stimuli (IAH) were then evaluated via a manikin task. Next, participants completed a picture-viewing task with simultaneous eye-tracking recording to assess viewing time of visual areas of interest (i.e., face and body). In our main hypothesis, we expected that in highly homophobic men, IAH would predict viewing time of pictures of homosexual couples, but not of heterosexual couples. Overall, the results of the linear mixed model analyses supported this hypothesis: homophobic men looked significantly longer at homosexual pictures only when they had a high IAH.

Table of Contents

1. Introduction ...... 1 1.1 Homophobia ...... 1 1.2 Dual-process models of behavior ...... 3 1.2.1 The impulsive or hot system ...... 3 1.2.2 The reflective or cool system ...... 4 1.2.3 Measuring impulsive and reflective precursors of behavior ...... 5 1.2.4 Interactions between the two systems ...... 6 1.2.5 The present study ...... 7

2. Method ...... 8 2.1 Participants ...... 8 2.2 Materials and Procedure ...... 9 2.2.1 ...... 9 2.2.2 Modern Homonegativity Scale (MHS) ...... 9 2.2.3 Manikin Task...... 10 2.2.4 Picture Viewing Task ...... 12 2.3 Data analyses ...... 13

3. Results ...... 13 3.1 Descriptive Statistics ...... 13 3.2 Results ...... 14 3.2.1 Results of viewing time analyses ...... 14 3.2.2. Results of analyses on explicit evaluation ...... 16

4. Discussion...... 18

5. References ...... 23

6. Appendices ...... 27

1

1. Introduction

1.1 Homophobia

In 1973 the American Psychiatric Association declassified as a mental disorder and removed it from its Diagnostic and Statistical Manual of Mental Disorders (DSM; Drescher, 2010). Just one year prior, Weinberg coined the term homophobia, which he defined as “the dread of being in close quarters with homosexuals – and in the case of homosexuals themselves, self- loathing” (as cited in Cullen & Barnes-Holmes, 2008). These two events reflect the beginning of a shift, not only within the mental health community, but a societal one as well, with a general trend of increasing liberalization of explicit attitudes towards homosexuality since the 1970s (Cullen & Barnes-Holmes, 2008). As homosexuality became more accepted by the mainstream, it was no longer the homosexuals themselves who were the problem, but rather those who feared them. The concept defined by Weinberg has spurred extensive research over the years, with over 2,700 citations returned by a search of the term “homophobia” in the PsychInfo database as of October 2015. However, the term is not consistently defined from study to study, with many studies including negative attitudes toward homosexuality in their definition, a usage criticized by those who argue that these attitudes do not necessarily arise from fear (Rye & Meaney, 2010; Smith, Oades, & McCarthy, 2012), but may, for example, result from strong religious or political beliefs (Herek, 2004). Others have questioned the validity of homophobia as an actual phobia, generally defined as a strong or irrational fear of something, and clinically defined in the most current edition of the DSM as an intense fear or anxiety about a particular object or situation (American Psychiatric Association, 2013). As a result of these objections, alternative terms have been proposed, such as homonegativity, , and sexual prejudice. In particular, Hudson and Ricketts (1980) recommended using homonegativity to describe negative reactions to homosexuality in general, while reserving the term homophobia to those cases in which individuals experience more irrational, emotional responses such as “fear, disgust, anger, discomfort, and aversion”. Despite these attempts at differentiation, literature searches of homophobia and associated terms have found no clear consensus on the exact definition of individual terms and a similarity between terms which can render them practically indistinguishable (Lottes & Grollman, 2010; Smith, Oades, & McCarthy, 2012).

2

Besides the uncertain terminology, the underlying causes of homophobia are also unclear. One possible explanation offered by psychoanalytic theorists posits that homophobia arises from repressed homosexual urges or from latent homosexuality, a homosexual attraction of which the individual is unaware or denies (Adams, Wright, & Lohr, 1996). More specifically, homophobia can represent a reaction formation to conscious or unconscious feelings of homosexual attraction. Reaction formation is a type of defense mechanism in which feelings or impulses considered unacceptable by the individual are masked by behaviors or values that run opposite to them (Weinstein et al., 2012). In other words, individuals who experience homosexual feelings which are incompatible with their idealized view of themselves may react to these feelings with homophobic behavior. A 1996 study by Adams, Wright, and Lohr set out to test this theory, by investigating the link between homophobia and sexual arousal to homosexual stimuli. In this study, 64 heterosexual, male university students were divided into two groups, homophobic or nonhomophobic, based on their scores on the Index of Homophobia scale (Hudson & Ricketts, 1980). Participants watched 4-minute segments of 3 different types of explicit erotic films depicting heterosexual activity, male homosexual activity, and female homosexual activity, during which changes in their penile circumference were measured using penile plethysmography. The results showed that homophobic men, but not nonhomophobic men, became physiologically aroused (i.e. experienced a significant increase in penile circumference) when watching the male homosexual videos. Furthermore, homophobic men were more likely to underestimate their own arousal in subjective evaluations. A possible interpretation, according to the authors, is that the arousal experienced by homophobic men is an expression of an unconscious or denied attraction to the same sex. By their own admission, this is not the only possible explanation: a previous study by Barlow, Sakheim, and Beck (1983) found that increases in penile circumference while watching erotic films can also be linked to anxiety. Therefore, the homophobic men in Adams et al.’s study may have not been sexually aroused by the homosexual stimuli, but may have instead been anxious. Another possibility is that homophobic men do not comprise a homogenous group. In fact, in the Adams et al. study (1996), only about half (54%) of the homophobic men showed “definite tumescence” (i.e., increase in penile circumference) to the homosexual video, while the other half showed only moderate (26%) or no significant (20%) tumescence. Other studies support the idea that there may more than one type of homophobic man. A 1984 series of studies by Shields and Harriman found a defensive response heart rate pattern, which is normally elicited by fear-provoking stimuli, in some, but not all highly

3 homonegative men while looking at pictures of explicit sexual activity between two men. A more recent study by Meier, Robinson, Gaither and Heinert (2006) used a picture viewing task, in which participants could view images of gay and heterosexual couples for as long as they liked, and found that homophobic men spent less time viewing homosexual than heterosexual pictures. A closer analysis found that this link was significant only in the homophobic men who were also high in self-deception1. Both studies, while failing to find the association between homophobia and same sex-attraction uncovered by Adams et al. (1996), suggest an interindividual variability within responses of homophobic men. That is to say, if the link between homophobia and an unconscious same-sex attraction exists, it may not necessarily exist in all homophobic individuals.

1.2 Dual-process models of behavior

To account for this interindividual variability, dual-process models, such as Metcalfe and Mischel’s (1999) hot and cool systems, or Strack and Deutsch’s (2004) reflective- impulsive model (RIM) seem particularly fitting. These models can help us understand why an individual’s behavior may not always reflect his beliefs or goals, as is the case when a homophobic man is sexually aroused by watching homosexual videos. According to these models, our behavior is not only determined by our reasoned, deliberate convictions, but also by our automatic, impulsive urges, which are guided by a reflective/cool system and an impulsive/hot system, respectively.

1.2.1 The impulsive or hot system

This system, known as the impulsive system in Strack and Deutsch’s (2004) model and as the hot system in Metcalfe and Mischel’s (1999) model, is responsible for impulsive, automatic behavior. It is emotionally-driven and responsible for quick approach-avoidance reactions to stimuli (Metcalfe & Mischel, 1999). In the RIM, the impulsive system is made up of associative clusters in long-term memory that can be activated by a perceived or imagined stimulus (Strack & Deutsch, 2004; Hofmann, Friese, & Strack, 2009). The strength of the links between clusters is relatively stable and changes only gradually though learning and experience (Strack & Deutsch, 2004). More specifically, these clusters can be created or strengthened when external stimuli are presented in close temporal or spatial proximity with

1 Measured using Paulhus’s Balanced Inventory of Desirable Responding, a 20-item scale tapping “unconscious tendencies to deceive the self through self-enhancement and denial" (as cited by Meier et al., 2006).

4 cognitive, affective, or motor reactions (Strack & Deutsch, 2004). For example, through repeated experience with homosexuality an associative cluster may be formed linking the concept of homosexuality with positive or negative affect experienced by the individual, and with the behavioral schema that led to this affect (e.g., hugging a member of the same sex). The formation of these clusters allows the individual to respond quickly when encountering comparable situations in the future, in a way that corresponds to his needs and previous experiences (Hofmann et al., 2009). For instance, once a “homosexuality” cluster is established, future encounters with homosexual stimuli can reactivate the cluster, automatically triggering an impulse in which a hedonic value (positive or negative) is assigned and a behavioral schema (approach or avoidance) is elicited in response to the stimulus (Hofmann et al., 2009). According to Strack and Deutsch (2004), approach reactions are linked to positive affect and to the processing of positive information, while avoidance reactions are linked to negative affect and the processing of negative information.

1.2.2 The reflective or cool system

This system, known as the reflective system in Strack and Deutsch’s (2004) model and as the cool system in Metcalfe and Mischel’s (1999) model, is responsible for controlled, deliberate behavior. It is cognitively-driven and knowledge-based, and is responsible for slow, rational, reflective, and strategic reactions (Metcalfe & Mischel, 1999). Rather than automatic approach-avoidance patterns, this system consists of “verbal or non-verbal descriptions, statements, assertions and commentaries” (Metcalfe & Mischel, 1999). In this system, a preferred behavioral option is selected following a decision process in which the value and probability of potential consequences is considered (Strack & Deutsch, 2004). For example, if an individual finds himself in a situation (e.g. watching a film in which there is a sex scene between two men) that contradicts his values (e.g. “I don’t like watching men having sex”), he may make a decision to override the unwanted behavior (“I will stop looking at this”) and activate a corresponding behavioral schema (by looking away from the screen) (Hofmann et al., 2009). Because of its reflective rather than reflexive nature, this system allows for complex, higher order mental operations that allow the individual to overcome stimulus control by exercising self-control or “willpower” over his impulses, as well as the flexibility to perform a variety of tasks (Hofmann et al., 2009; Metcalfe & Mischel, 1999). This comes at a cost, however, with speed being one of the trade-offs. The reflective system is not only slower than

5 the impulsive system, but also requires considerable cognitive resources, which means its operation can be hindered when, for example, the individual is distracted or has very high or low levels of arousal (Strack & Deutsch, 2004).

1.2.3 Measuring impulsive and reflective precursors of behavior

As the two systems are guided by different processes, it follows that they will be measured in different ways. Hofmann et al. (2009) lay out 4 criteria for measuring impulsive precursors of behavior. Firstly, the measure should be sufficiently specific to the temptation in question. For example, if we want to predict behavior towards homosexual men, we should measure impulses towards men who are clearly homosexual, rather than towards men in general. Second, the measure should capture spontaneously activated reactions to stimuli which tap into underlying the associative structures of impulsive processing. Specifically, instead of asking participants directly about their impulses, implicit measures are used. Third, since impulses are thought to be triggered automatically, the measure should avoid the contaminating influence of conscious control. Fourth, the measure should be sensitive to individual variation in impulse strength, be it dispositional or situational. Implicit measures, in which the impact of the psychological construct measured on participant response is unintentional, independent of available resources, unconscious or uncontrollable, (Gawronski & De Houwer, 2014) seem well-suited to measure the impulsive precursors of behavior. The Implicit Association Test (IAT) (Greenwald, McGhee, & Schwartz, 1998) measures implicit attitudes by measuring automatic affective reactions, while other measures, like Krieglmeyer and Deutsch’s (2010) Manikin task capture the behavioral aspect of an impulse (i.e., approach-avoidance reactions) (Hofmann, et al., 2009). On the other hand, given that the reflective precursors of behavior are based upon declarative propositions that can be communicated to others, explicit measures may be used. Contrary to implicit measures, explicit measures are those in which the impact of the psychological construct measured is intentional, dependent on available resources, conscious, or controllable (Gawronski & De Houwer, 2014). Using verbal self-reports (e.g., questionnaires), reflective precursors of behavior, such as values, beliefs, and deliberate evaluations, can be properly measured.

6

1.2.4 Interactions between the two systems

The hot/impulsive system and cool/reflective system are not completely independent, but run in parallel to each other and are in continuous interaction (Mischel & Ayduk, 2004; Strack & Deutsch, 2004). In the RIM, both systems can activate behavioral schemata that can result in overt behavior when an activation threshold is reached (Strack & Deutsch, 2004). Several schemata can be activated at the same time, which may result in conflict if they are incompatible and inhibit one another (Strack & Deutsch, 2004). For instance, if we take the previous example of an individual who doesn’t enjoy watching a film in which there is a homoerotic scene, his reflective system, in response to the discrepancy between the situation and personal values, may activate the schema “stop looking at this”. If this individual has an impulsive avoidance towards homosexual stimuli, the same behavioral schema will also be activated by the impulsive system, and the execution of the behavior (i.e., to look away from the screen) will be facilitated (Strack & Deutsch, 2004). However, if this individual instead has an impulsive attraction to homosexuals, his impulsive system may activate the schema “keep looking at this”, which is incompatible with the schema activated by the reflective system. This incompatible activation can be accompanied by a feeling of internal conflict and temptation (Hofmann et al., 2009; Strack & Deusch, 2004). In a self-control conflict, the behavioral schema that “wins” (i.e., that determines overt behavior) depends on the relative strength of activation of competing schema, which can be moderated by a number of factors (Hofmann et al., 2009). As previously mentioned, the reflective system is limited by its dependence on cognitive resources, and if these cognitive resources are compromised (e.g., in situations of high cognitive load or after alcohol consumption), the reflective system may not succeed in sufficiently activating its corresponding schema, and the schemas triggered by the impulsive system become more likely to influence overt behavior (Hofmann et al., 2009). Furthermore, dispositional factors, such as working memory capacity (Hofmann, Gschwendner, Friese, Wiers, & Schmitt, 2008), trait self-control (Friese & Hofmann, 2009), and ability to control attention (Friese & Hofmann, 2012) may also moderate the relative strengths of the two systems. Prior experimental studies have shown that people vary, not only in their explicit attitudes, but also in their impulsive tendencies (Friese & Hofmann, 2009). Additionally, regarding impulsive and reflective reactions toward homosexuals, a recent study by MacInnis

7 and Hodson (2013) found no link between implicit homosexual attraction2 and explicit attitudes about homosexuals3 among men, and only a weak correlation among women. This dissociation between implicit and explicit attitudes underlines the importance of considering impulsive measures in addition to reflective measures to measure homosexual interest.

1.2.5 The present study

While Adams et al. (1996) took into account variability in reflective attitudes (i.e., by measuring levels of homophobia), they did not consider that individuals may also vary in their impulses. Consequently, their study implies that all homophobic men have an unconscious sexual attraction toward men. However, based on dual-process models, it is possible that only a portion of these men are implicitly attracted to homosexual stimuli, while the others either have no attraction to these stimuli, or may even be repulsed by them. Moreover, differences in impulsive processes may explain why some of the homophobic men in Adams et al.’s (1996) study showed no increase in penis circumference in response to homosexual stimuli. The goal of the present study is therefore to determine if the link between homophobia and homosexual attraction is moderated by impulsive approach- avoidance tendencies toward homosexual stimuli. More specifically, we want to know if homophobic men who show an impulsive attraction to homosexual stimuli also show sexual interest in viewing photographs of homosexual couples. As previously mentioned, Adams et al. (1996) measured sexual interest in their study using penile plethysmography, a method not without its share of limitations, including its reliance on expensive equipment, its need for a motivated and physiologically responsive subject, and its highly invasive nature (Laws & Gress, 2004). Instead we opted to use a picture viewing task, in which the participant chooses how long he wishes to view a stimulus, which has been previously shown to reliably indicate sexual interest (Friese & Hofmann, 2012; Gress, 2005). Specifically, it has been shown that participants look longer at images depicting their preferred sexual object (e.g. a heterosexual man will look longer at pictures of nude women vs. nude men) (Laws & Gress, 2004). Moreover, whereas increases in penile circumference may be interpreted as anxiety (Barlow et al., 1983), a prior viewing task study found that phobic individuals look less long at images which they consider threatening (Tolin, Lohr, & Sawchuk, 1999). A picture viewing task may therefore provide for a less

2 As measured by the Sexual Preference IAT, validated by Snowden, Wichter, and Gray (2008), an implicit measure of sexual orientation that has been found to significantly correlate with explicit measures of sexual orientation. 3As measured by the Index of Homophobia scale (Hudson & Ricketts, 1980)

8 ambiguous interpretation of results. Additionally, this task used eye tracking technology to track participants’ eye gaze. This allowed us to exclusively measure the time spent looking at the relevant characteristics of the images (i.e., face and body). In the present study, homonegative attitudes were measured via a questionnaire administered in a group setting. This was followed by an individual visit to our lab, where participants first completed a task designed to measure their impulsive tendencies to approach or to avoid homosexual stimuli. Finally, participants completed the picture viewing task. During this task, participants were also asked to rate each picture on its pleasantness. This explicit evaluation allowed us to test if the effect of homophobia on same-sex interest could extend to measures under conscious control, or if it is only limited to largely uncontrollable measures (i.e., viewing time). As a first hypothesis, we expected that in participants with a high level of homonegativity, an impulsive tendency to approach homosexual stimuli (IAH) would increase the time spent viewing pictures of homosexual couples in the picture viewing task, as a result of a conflict between reflective/cool processes (i.e., negative attitudes towards homosexuals) and impulsive/hot processes (i.e., IAH). Whereas there would be no effect of IAH on viewing time of heterosexual images, since looking at these pictures would not necessarily trigger a conflict between reflective and impulsive processes in heterosexual participants. A second hypothesis is that in participants with a high level of homonegativity, IAH would have no effect on the explicit evaluation of images, whether heterosexual or homosexual. Indeed, unlike viewing time, explicit evaluation is, in principle, under complete conscious control and therefore would not be influenced by impulsive processes.

2. Method

2.1 Participants

Forty-eight male students were recruited from the Psychology Department Research Subject Pool at the University of Geneva. All had normal or corrected vision. Three participants who identified as homosexual were excluded from analyses. Seven additional students were excluded from analyses when a post-session debriefing revealed they were aware of the variable measured by the picture viewing task. Analyses were run on the remaining 38 heterosexual participants (Mage = 22 years, SD= 4.74), who participated in the

9 study in exchange for course credit. All participants were given written informed consent prior to participation, and received a written debriefing at the end of each study session.

2.2 Materials and Procedure

Participants first completed the Modern Homonegativity Scale (Morrison & Morrison, 2002; See Appendix A) during large group testing at the start of the semester, as well as a question determining their sexual orientation. One month later, participants came to the laboratory individually and were seated in front of a computer in order to complete a manikin task and a picture-viewing task. During the picture-viewing task, they were seated at a 75-cm distance from the computer screen. At this distance, eye tracking product work optimally (Harrison, Gray, & Critchley, 2009; Harrison, Singer, Rotshtein, Dolan, & Critchley, 2006; Lavin, San Martin, & Rosales Jubal, 2014). Participants begin with a nine-point eye tracker calibration, before the start of the first trials of the picture-viewing task.

2.2.1 Sexual Orientation

Participants were asked the following question concerning sexual orientation: “I consider my sexual orientation to be (choose only one)”. Participants had four response options: 1 = straight (heterosexual); 2= bisexual; 3 = gay; 4= asexual (see Meier et al., 2006 for similar procedure). In our study, only participants who gave the first response were retained for analyses.

2.2.2 Modern Homonegativity Scale (MHS)

Homonegative attitudes were assessed by the 12-item Modern Homonegativity scale (MHS; Morrison & Morrison, 2002). This scale was created with the goal of measuring more contemporary, subtle attitudes towards homosexuals such as those held by university students (Morrison & Morrison, 2002). It correlates strongly with the Index of Homophobia (Hudson & Ricketts, 1980) scale used by Adams et al. (1996), while having the advantage of resulting in slightly more normally distributed data (Rye & Meaney, 2010). In this study, the internal consistency of the scale was good” (alpha =.85). The items (e.g., “ should stop shoving their lifestyle down other people’s throats”; “The notion of universities providing students with undergraduate degrees in Gay Studies is ridiculous”) were rated on a 5-point scale (1 = strongly disagree; 5 = strongly agree). Item scores were averaged; higher scores indicate more negative attitudes toward gay men.

10

2.2.3 Manikin Task

The Manikin task (Krieglmeyer & Deutsch, 2010) was used to assess participants’ spontaneous approach-avoidance tendencies toward gay men. In this task, participants were asked move a schematic image of a human figure (i.e., a “manikin”) on the computer screen upwards or downwards by pressing the “8” key (up) or the “2” key (down) with their middle finger as quickly and as accurately as possible. Each trial started with a fixation cross at the center of the screen. Using their middle finger, participants had to press the “5” key when they were ready to begin the trial, and keep it pressed until they were ready to move the manikin. Triggered by the key press, a manikin then appeared either in the upper or lower half of the screen with equal probability. Then, 750ms after its appearance, a black and white image of a heterosexual or homosexual couple (see Figure 1 for examples) was presented at the center of the screen. Five hundred ms after the response key press, the screen was cleared for 1,000 ms before the start of the next trial. All 10 images (5 homosexual, 5 heterosexual) were downloaded from publicly available websites (see Appendix B).

11

Figure 1. Sample trials from the Manikin task. A & B represent the condition where participants were instructed to approach (A) heterosexual stimuli and to avoid (B) homosexual stimuli. C & D represent the condition where participants were instructed to avoid (C) heterosexual stimuli and approach (D) homosexual stimuli. The up and down arrows represent the keyboard key which the participant had to press for the right answer. Adapted from Cheval, Sarrazin, & Pelletier (2014).

Each participant completed 2 trial blocks, representing the two conditions of the task. In one condition, the participants were asked to move the manikin as quickly and as accurately as possible “away” from homosexual images and “toward” heterosexual images. In the other condition, the instructions were reversed: participants were asked to approach homosexual images and avoid heterosexual images. If an incorrect response was made, error feedback appeared on the screen. Each block consisted of 12 practice trials and 40 test trials, so that each of the 10 images were shown with the manikin twice at the top and twice at the bottom of the screen. The order of the blocks was counterbalances across participants.

12

Before analyzing the data we excluded incorrect responses (10.44%) as well as responses below 150 ms (0.10%) and above 1,500 ms (2.91%) as recommended by Krieglmeyer & Deutsch (2010). Results are presented with the median reaction time (RT). Participants’ impulsive approach tendencies toward gay men were calculated by subtracting the median approach RT from the median avoidance RT toward homosexual images. A positive score always indicates a tendency to approach rather than avoid homosexual stimuli.

2.2.4 Picture Viewing Task

As the final task, participants were asked to look at and evaluate 25 images, rating them on how pleasant or unpleasant they find them, under the pretense of helping us choose images for use in a future study. Ten images depicted homosexual couples, 10 depicted heterosexual couples, and 5 were neutral. Two of the neutral images served as practice trials. Neutral images4 were taken from the International Affective Picture System (Lang, Bradley, & Cuthbert, 2005). All heterosexual and homosexual images were downloaded from publicly available sites. Half of the homosexual and heterosexual images were previously used in the manikin task, while the other half were novel. To control visual features across images, all images were presented in black in white and at the same size (200 x 500 pixels). Content in the heterosexual and homosexual images was also matched. Participants were instructed to view the picture as long as needed in order to give it an accurate rating. They were told to then press the space bar when they were ready to evaluate the image. Then, a rating scale appeared on the screen ranging from 1 (very unpleasant) to 9 (very pleasant), and participants were told to press a number on the keyboard to rate the image. After the number was pressed, a blank screen appeared for 500ms before the next picture was shown (see Meier et al., 2006 for similar procedure). The rating of the first two practice images were completed with the experimenter present, to make sure that the task instructions were understood. The experimenter then left the room and the participant was left alone for the rest of the task, as prior research has shown that the presence of an audience can influence viewing times and ratings of sexual material (Yuen & Ickes, 1984). Two dependent measures were used in the picture-viewing task: (a) the explicit rating of the picture and (b) the viewing time of homosexual and heterosexual images. Viewing

4 Neutral images were included due to a previous hypothesis concerning pupil dilation. This hypothesis does not appear in this paper due to the complexity of the statistical analyses required to interpret the data. As a result, the viewing time and evaluation data for neutral images were excluded from our analyses.

13 time was recorded using the Tobii T60 eye tracker (Tobii Technology, Inc., Danderyd, Sweden) and Tobii Studio Version 2.3.2 (Psychology Software Tools, Inc., http://www.pstnet.com/hardware.cfm?ID=107). Areas of interest were delimited in order to separate the photograph subjects from the background, and the total time spent looking at the people (i.e., face and body) in the images5 was used in our analyses as the viewing time variable.

2.3 Data analyses

First, we used descriptive statistics and bivariate correlation between IAH and homonegativity. Then, the viewing time and the explicit evaluation of homosexual and heterosexual photographs were analyzed using Linear Mixed Models (LMMs). Traditional analyses (e.g., ANOVA) disregard the sampling variation of stimuli, despite numerous warnings about the shortcomings of this practice (Clark, 1973; Judd, Westfall, & Kenny, 2012). Indeed, treating only participants as a random effect increases the likelihood of Type I errors (Judd et al., 2012). LMMs allow correct estimations of the parameters for analyzing data with multiple crossed-random effects (e.g., participants and stimuli). Consequently, the results obtained from crossed- random LMMs can be generalized to both participants and stimuli, and are much less vulnerable to Type-1 errors (Judd et al., 2012). In these analyses, IAH, explicit level of homophobia, and the type of photographs, as well as their interactions were included as predictors in the models. In all analyses, the independent variables were centered in the case of continuous variables, and dummy coded in the case of dichotomous variables (i.e., old vs. new and homosexual vs. heterosexual images). After having run the LMMs, we calculated a pseudo R2 to estimate by how much the inclusion of predictors in the model would decrease the error rate.

3. Results

3.1 Descriptive Statistics

Table 1 summarizes the characteristics of the independent variables, which show that in general, the participants had a moderate level of homonegativity (m = 2.60) and a tendency to approach homosexual stimuli nearly equivalent to the tendency to avoid them (m = -2.6ms, SD = 122.54).

5 A prior analysis considered the viewing time of the pictures as a whole and found similar results.

14

The independent variables were well-distributed and did not correlate significantly (r = -.224, p = .177). These results suggest that there is a variability in the participants’ IAH regardless of their homonegativity level. (Appendix C).

Table 1 Characteristics of variables of interest Variables M SD Min Max MHS 2.605 .736 1.167 3.667 IAH -2.658 122.543 -266 218 Note. N = 38. MHS = Modern Homonegativity Scale (Morrison & Morrison, 2002). IAH = Impulsive tendency to approach homosexual stimuli.

3.2 Results

3.2.1 Results of viewing time analyses

The results of the linear mixed model analysis of viewing time revealed a significant three-way interaction between homonegativity, IAH, and images (b = 2.353, p = .016), indicating that the effect of IAH on viewing time significantly varies depending on the type of image (heterosexual vs. homosexual) between high homonegativity and low homonegativity participants (Table 2). For participants with a high score of homonegativity, results revealed that the effect of IAH on viewing time significantly varies depending on image type (b = 2.415, p = .008). Specifically, simple slopes analyses (Figure 2a) showed that IAH is positively related to viewing time of homosexual images (b = 4.076, p = .028), but not of heterosexual images (b = 1.660 , p = .360). Simple effects tests further revealed that participants with a high homonegativity score looked significantly longer at homosexual images than heterosexual images when they had a high IAH (i.e., +1 SD) (b = -362.953, p = .048, while participants with a low IAH (i..e., -1 SD) did not look longer at homosexual images (b = 227.806, p = .131). In contrast, for participants with a low score of homonegativity, results revealed that the effect of IAH on viewing time does not significantly vary depending on image type (b = - 0.926, p = .340). Specifically, simple slopes analyses (Figure 2b) showed that IAH is not

15 significantly related to viewing time of either homosexual (b = 2.501, p = .196) or heterosexual (b = 3.428, p = .079) images. Simple effects tests further revealed that participants with a low score of homonegativity looked significantly longer at heterosexual images than homosexual images when they had a high IAH (i.e., +1 SD) (b = 401.607, p = .035), while participants with a low IAH (i.e., -1 SD) did not look longer at heterosexual images (b = 174.969, p = .255).

Figure 2. Viewing time of homosexual and heterosexual images as a function of the tendency to approach homosexual stimuli (IAH) for participants with a high (a) and low (b) level of homonegativity.

Furthermore, a negative main effect of homophobia emerged (b = -609.412, p = .009), indicating that on the whole, participants with a high score of homonegativity spent significantly less time looking at the images, regardless of type, compared to participants with a low score of homonegativity. A positive effect of old versus new images emerged (b =126.729, p = .013), indicating that participants tend to look longer at new, rather than old pictures. Finally, we have found that the mixed linear model explains 28.5% of the variance of participants’ viewing time.

16

3.2.2. Results of analyses on explicit evaluation

The linear mixed model analysis of the explicit evaluation of images revealed a significant three-way interaction between homonegativity, IAH, and images (b = 0.004, p = .007), indicating that the effect of IAH on explicit evaluation significantly varies depending on the type of image (heterosexual vs. homosexual) between high homonegativity and low homonegativity participants (Table 2). For participants with a high score of homonegativity, results revealed that the effect of IAH on explicit evaluation significantly varies depending on image type (b = 0.009, p < .001). Specifically, simple slopes analyses (Figure 3a) showed that IAH is positively related to explicit evaluation of homosexual images (b = 0.008, p = .001), but not of heterosexual images (b = -0.001, p = .676). Simple effects tests further revealed that participants with a high homonegativity score evaluate homosexual images as significantly less pleasant than heterosexual images when they had a low IAH (i.e., -1 SD) (b = 4.110, p < .001), This difference remained significant but was reduced when participants had a high IAH (i.e., + 1ET) (b = 1.988, p < .001). For participants with a low score of homonegativity, results revealed that the effect of IAH on explicit evaluation significantly varies depending on image type (b = 0.004, p = .006). However, simple slopes analyses (Figure 3b) showed that IAH is not significantly related to the explicit evaluation of either homosexual (b = 0.002, p = .351) or heterosexual (b = -0.001, p = .493) images. Simple effects tests further revealed that participants with a low score of homonegativity evaluate heterosexual images more positively than homosexual images whether they had a high IAH (i.e., + 1 SD) (b = 1.878, p < .001) or a low IAH (i.e., - 1 SD) (b = 2.755, p < .001). Finally, we have found that the mixed linear model explains 12.5% of the variance of participants’ explicit evaluation.

17

Figure 3. Explicit evaluation of homosexual and heterosexual images as a function of the tendency to approach homosexual stimuli (IAH) for participants with a high (a) and low (b) level of homonegativity.

18

Table 2. General results of the linear mixed-model on viewing time and on explicit evaluation

Time viewing AOI Explicit evaluation Parameter b p b p Fixed effects Intercept 1756.254 < .001 5.874 < .001 Noveltya 126.729 .013 -.127 .113 MHS -609.412 .009 -.067 .791 IAH 2.916 .021 .002 .182 Imagesb -110.357 .248 -2.683 < .001 MHS * IAH -.068 .971 .002 .295 MHS * Images 250.635 .033 -.516 .001 IAH * Images .744 .246 .006 < .001 MHS * IAH * Images 2.353 .016 .004 .007 Random effects Residuals 1097549.10 < .001 2.016 < .001 Participants intercept 753414.867 < .001 .948 < .001 Images intercept 12573.918 .363 .063 .094 R2 .285 .125 Note. N = 38. MHS = Modern Homonegativity Scale (Morrison & Morrison, 2002). IAH = Impulsive approach tendency towards homosexual stimuli. AOI = areas of interest (in ms). aimages already presented in the manikin task = -0.5, new images = 0.5. b heterosexual images = -0.5, homosexual images = 0.5.

4. Discussion

In the present study, we investigated the role of impulses in same-sex attraction in homophobic individuals. Specifically, we sought to build upon Adams et al.’s (1996) finding that homophobic men were physiologically aroused by homosexual stimuli. According to the authors, and in line with psychodynamic theory, this result is evidence of an unconscious or denied attraction to the same sex in homophobic men. However, not all homophobic men in the study showed definite same-sex arousal. Dual-process models (Metcalfe & Mischel,

19

1999; Strack & Deutsch, 2004) could help account for this variability by characterizing same- sex attraction in homophobic individuals as a conflict between reflective and impulsive processes. Indeed, dual-process models suggest that individuals vary in their impulsive tendencies, and if these impulses are not measured, they can be misinterpreted as error variance (Friese & Hofmann, 2009). Some individuals may therefore be implicitly attracted to homosexual stimuli, while others may not. We thus set out to test if the impulsive tendency to approach homosexual stimuli could explain the variability in sexual attraction toward homosexual stimuli. Overall, the results support these assumptions. Firstly, impulsive tendencies were shown to vary among participants, and furthermore these tendencies were independent of homophobia level. Homophobic participants and non-homophobic participants were equally as likely to have either a low, moderate, or high tendency to approach homosexual stimuli. Accordingly, in some homophobic participants, there was a conflict between reflective and impulsive processes as described by dual-process models. In other words, these participants had a high impulsive tendency to approach homosexual stimuli in spite of their reflective negative attitudes towards homosexuals. In line with our first hypothesis, we found that these same participants looked longer at homosexual stimuli than homophobic participants with a low IAH. Additionally, they also looked longer at homosexual pictures than they looked at heterosexual pictures, unlike low IAH participants, who looked at both types of pictures about the same. Furthermore, by using eye-tracking technology, we were able to control that participants were in fact looking at the relevant parts of these pictures (i.e., face and body) rather than background elements. Regarding our second hypothesis, our findings were somewhat surprising. Because explicit evaluation falls under the conscious control of the participants, we had expected the results to reflect the participants’ reflective, rather than impulsive, processes. In other words, we anticipated that homophobic individuals would all rate the homosexual pictures similarly, irrespective of their impulsive tendencies. While all participants generally rated homosexual pictures as less pleasant than heterosexual pictures, homophobic participants with a high IAH rated homosexual pictures as more pleasant than homophobic participants with a low IAH. Therefore, despite the participants’ being able to control their responses in this task, impulsive tendencies still manage to influence the behavior of strongly homophobic individuals. A likely explanation for this unexpected result may be that the task of viewing pictures of homosexual couples induced an increased state of stress in these participants. In dual-process models, increased situational stress has been theorized to interfere with functioning of reflective processes, enabling impulsive processes

20 to guide behavior (Metcalfe & Mischel, 1999). It is therefore possible that upon viewing homosexual pictures, stress levels in homophobic participants were raised thereby compromising reflective processes, allowing automatic approach tendencies to influence the explicit evaluation of images. In spite of this, it is worth noting that responses were still largely under conscious control, with heterosexual images rated more positively than homosexual images in all cases. However, the percentage of explained variance is higher for viewing time than it is for explicit evaluation. Therefore, although impulsive processes may be able to influence behavior under conscious control, their impact is mainly bound to less controllable forms of behavior, in line with dual-process models. Our results therefore corroborate and expand upon Adam et al.’s (1996) work, supporting the finding that homophobic individuals may indeed have a higher sexual interest (Friese & Hofmann, 2012; Gress, 2005) toward homosexual stimuli over heterosexual stimuli, but only if they also have a strong impulse to approach homosexual stimuli. Taking into account inter-individual variability in impulsive tendencies as described by dual-process models may thus explain why only some men in Adam et al.’s (1996) study experienced physiological arousal to homosexual stimuli while others did not. It is possible that those men who showed no increase in penis circumference may have a low automatic approach tendency toward homosexual stimuli, while those who experienced a definite increase in penis circumference may have a strong approach tendency. The inclusion of impulsive tendencies as a moderating variable may also help clarify why other studies have failed to find consistent results on the association between homophobia and same-sex attraction (MacInnis & Hodson, 2013; Meier et al., 2006; Shields & Harriman, 1984). Moreover, unlike previous studies that have used viewing time as a measure of sexual interest, eye-tracking technology was utilized to restrict the measure of viewing time to relevant areas of the picture (i.e., face and body), thereby enhancing the measurement validity. Although participants with a low level of homophobia were not the focus of the study, analyses were also run on these individuals for the sake of comparison. In these men, automatic impulsive tendencies did not influence viewing time nor explicit evaluations for either type of picture. This finding is unsurprising, as it is consistent with the assumption of dual-process models that at low to moderate levels of stress, reflective processes are enhanced, and impulsive processes therefore have little impact on behavior (Mischel & Ayduk, 2004). Accordingly, previous studies have shown that viewing homosexual stimuli is not related to increases in anxiety or anger in non-homophobic men, in contrast with homophobic men (Bernat et al., 2001; Parrott & Zeichner, 2005). Whereas in homophobic

21 participants, anxiety and anger may weaken reflective processes, empowering impulsive tendencies, in non-homophobic participants, reflective processes may function as expected, resulting in behavior that corresponds to participants’ beliefs and reasoned attitudes. In comparing homophobic and non-homophobic individuals, results also revealed a main effect of homophobia. In other words, participants high in homophobia viewed all types of pictures, regardless of their nature (i.e., homosexual or heterosexual), for less time than participants low in homophobia did. Seeing that the Modern Homonegativity Scale used in our study is positively correlated to political conservatism and religiosity (Morrison & Morrison, 2002), this result may reflect a divergence between the sexual nature of the photographs and conservative religious or political beliefs potentially held by homophobic men. These men may feel uncomfortable viewing these types of images, and consequently would spend less time looking at them. More research is needed, however, to ascertain the veracity of this interpretation. A few limitations of the present study should be mentioned. First, the sample was entirely composed of male university students. As a result, it is difficult to say whether our findings can be generalized to a larger population. Future research should draw upon a more heterogenous population in order to increase the external validity of these results. Second, the present study was carried out under laboratory conditions which may not adequately generalize to real world conditions involving homophobia. Future studies should focus on more ecologically valid ways to measure sexual interest. Future studies should also take into account other potential moderators that can affect the reflective-impulsive balance. As previously mentioned, at low levels, stress can enhance reflective processes and, at high levels, weaken them (Mischel & Ayduk, 2004). The amount of stress an individual experiences at any given moment is dictated by both their chronic levels of stress and stress elicited by the current situation. Measuring chronic and situational stress could help determine their impact on the link between homophobia and same-sex attraction. Additionally, two previous studies involving the viewing time of sexual stimuli have found that working memory (Hofmann et al., 2008) and attention control (Friese & Hofmann, 2012) both play a role in how impulsive and reflective processes influence behavior. Impulsive processes were most likely to predict viewing time of erotic material when men had a low working memory capacity (Hofmann et al., 2008) or when they were low in attention control (Friese & Hofmann, 2012). Controlling for these variables in future studies would provide a more complete picture of our findings.

22

In conclusion, the present study investigated the contribution of impulsive processes in the behavior of homophobic men. Prior studies have found mixed results concerning the association between homophobia and attraction to homosexual stimuli. Specifically, it appears that some, but not all, homophobic men hold an attraction to the same sex. Our findings suggest that this inconsistency can be explained by an inter-individual variability in impulsive tendencies. These results highlight the need to consider both reasoned attitudes and automatic impulses in predicting sexual-related behaviors.

23

5. References

Adams, H. E., Wright, L. W., & Lohr, B. A. (1996). Is homophobia associated with homosexual arousal? Journal of abnormal psychology, 105 (3), 440•445. doi:10.1037/0021•843X.105.3.440 American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. doi:10.1176/appi.books.9780890425596 Barlow, D. H., Sakheim, D. K., & Beck, J. G. (1983). Anxiety increases sexual arousal. Journal of Abnormal Psychology, 92(1), 49. doi:10.1037/0021-843X.92.1.49 Bernat, J. A., Calhoun, K. S., Adams, H. E., & Zeichner, A. (2001). Homophobia and physical aggression toward homosexual and heterosexual individuals. Journal of Abnormal Psychology, 110(1), 179. doi: 10.1037/0021-843X.110.1.179 Cheval, B., Sarrazin, P., & Pelletier, L. (2014). Impulsive approach tendencies towards physical activity and sedentary behaviors, but not reflective intentions, prospectively predict non-exercise activity thermogenesis. Plos one, 9(12), e115238. doi:10.1371/journal.pone.0115238 Clark, H. H. (1973). The language-as-fixed-effect fallacy: A critique of language statistics in psychological research. Journal of verbal learning and verbal behavior, 12(4), 335- 359. doi: 10.1016/S0022-5371(73)80014-3 Cullen, C., & Barnes-Holmes, D. (2008). Implicit pride and prejudice: A heterosexual phenomenon? In T. G. Morrison & M. A. Morrison (Eds.), Modern Prejudice. New York: Nova Science. Drescher, J. (2010). diagnoses: Parallels and contrasts in the history of homosexuality, gender variance, and the Diagnostic and Statistical Manual. Archives of sexual behavior, 39(2), 427-460. doi: 10.1007/s10508-009-9531-5. Friese, M., & Hofmann, W. (2009). Control me or I will control you: Impulses, trait self•control, and the guidance of behavior. Journal of research in personality, 43 (9). doi:10.1016/j.jrp.2009.07.004 Friese, M., & Hofmann, W. (2012). Just a little bit longer: Viewing time of erotic material from a self•control perspective. Applied cognitive psychology, 26 (3), 489•496. doi:10.1002/acp.2831 Gawronski, B., & De Houwer, J. (2014). Implicit measures in social and personality psychology. In H. T. Reis & C. M. Judd (Eds.), Handbook of research methods in

24

social and personality psychology (pp. 283•310). New York, NY: Cambridge University Press. doi: 10.1017/CBO9780511996481.016 Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. (1998). Measuring individual differences in implicit cognition: the implicit association test. Journal of personality and social psychology, 74(6), 1464. doi: 0.1037/0022-3514.74.6.1464 Gress, C. L. Z. (2005). Viewing time measures and sexual interest: another piece of the puzzle. Journal of sexual agression, 11 (2), 117•125. doi:10.1080/13552600500063666 Harrison, N. A., Gray, M. A., & Critchley, H. D. (2009). Dynamic pupillary exchange engages brain regions encoding social salience. [Research Support, Non-U.S. Gov't]. Soc Neurosci, 4(3), 233-243. doi:10.1080/17470910802553508 Harrison, N. A., Singer, T., Rotshtein, P., Dolan, R. J., & Critchley, H. D. (2006). Pupillary contagion: central mechanisms engaged in sadness processing. Social Cognitive and Affective Neuroscience, 1(1), 5-17. doi:10.1093/scan/nsl006 Herek, G. M. (2004). Beyond “homophobia”: Thinking about sexual prejudice and stigma in the twenty-first century. Sexuality Research and Social Policy,1(2), 6-24. doi:10.1525/srsp.2004.1.2.6 Hofmann, W., Friese, M., & Strack, F. (2009). Impulse and self•control from a dual•system perspective. Perspectives on psychological science, 4 (2), 162•176. doi:10.1111/j.1745•6924.2009.01116.x Hofmann, W., Gschwendner, T., Friese, M., Wiers, R. W., & Schmitt, M. (2008). Working memory capacity and self•regulatory behavior: toward an individual differences perspective on behavior determination by automatic versus controlled processes. Journal of personality and social psychology, 95 (4), 962•977. doi:10.1037/a0012705 Hudson, W. W., & Ricketts, W. A. (1980). A strategy for the measurement of homophobia. Journal of homosexuality, 5 (4), 357•372. doi:10.1300/J082v05n04_02 Judd, C. M., Westfall, J., & Kenny, D. A. (2012). Treating stimuli as a random factor in social psychology: a new and comprehensive solution to a pervasive but largely ignored problem. Journal of personality and social psychology, 103 (1), 54•69. doi:10.1037/a0028347 Krieglmeyer, R., & Deutsch, R. (2010). Comparing measures of approach–avoidance behaviour: The manikin task vs. two versions of the joystick task. Cognition and Emotion, 24(5), 810-828. doi:10.3389/fpsyg.2014.01398

25

Lang, P., Bradley, M. & Cuthbert, B. (2005). International affective picture system (IAPS): Affective ratings of pictures and instruction manual (Technical Report A-6). Gainesville, FL: University of Florida. Lavin, C., San Martin, R., & Rosales Jubal, E. (2014). Pupil dilation signals uncertainty and surprise in a learning gambling task. Front Behav Neurosci, 7, 218. doi:10.3389/fnbeh.2013.00218 Laws, D. R., & Gress, C. L. (2004). Seeing things differently: The viewing time alternative to penile plethysmography . Legal and criminological psychology, 9 (2), 183•196. doi:10.1348/1355325041719338 Lottes, I. L., & Grollman, E. A. (2010). Conceptualization and assessment of homonegativity. International Journal of Sexual Health, 22(4), 219-233. doi:10.1080/19317611.2010.489358 MacInnis, C. C., & Hodson, G. (2013). Is homophobia associated with an implicit same-sex attraction?. Journal of sex research, 50(8), 777-785. doi:10.1080/00224499.2012.690111 Meier, B. P., Robinson, M. D., Gaither, G. A., & Heinert, N. J. (2006). A secret attraction or defensive loathing? Homophobia, defense, and implicit cognition. Journal of Research in Personality, 40(4), 377-394. doi:10.1016/j.jrp.2005.01.007 Metcalfe, J., & Mischel, W. (1999). A hot/cool-system analysis of delay of gratification: dynamics of willpower. Psychological review, 106(1), 3. doi:10.1037/0033- 295X.106.1.3 Mischel, W., & Ayduk, O. (2004). Willpower in a cognitive•affective processing system: the dynamics of delay of gratification. In R. F. Baumeister & K. D. Vohs (Eds.), Handbook of Self•Regulation: Research, Theory, and Applications, (pp. 83•105). New York, NY: Guildford Press. Morrison, M. A., & Morrison, T. G. (2002). Development and validation of a scale measuring modern prejudice toward gay men and women. Journal of homosexuality, 43(2), 15-37. doi:10.1300/J082v43n02_02 Parrott, D. J., & Zeichner, A. (2005). Effects of sexual prejudice and anger on physical aggression toward gay and heterosexual men. Psychology of Men & Masculinity, 6(1), 3. doi:10.1037/1524-9220.6.1.3

26

Rye, B. J., & Meaney, G. J. (2010). Measuring homonegativity: A psychometric analysis. Canadian Journal of Behavioural Science/Revue canadienne des sciences du comportement, 42(3), 158. doi:10.1037/a0018237 Shields, S. A., & Harriman, R. E. (1984). Fear of male homosexuality: Cardiac responses of low and high homonegative males. Journal of Homosexuality,10(1-2), 53-67. doi:10.1300/J082v10n01_04 Smith, I. P., Oades, L., & McCarthy, G. (2012). Homophobia to heterosexism: Constructs in need of re-visitation. Gay and Lesbian Issues and Psychology Review, 8(1), 34–44 Snowden, R. J., Wichter, J., & Gray, N. S. (2008). Implicit and explicit measurements of sexual preference in gay and heterosexual men: A comparison of priming techniques and the implicit association task. Archives of sexual behavior, 37(4), 558-565. doi:10.1007/s10508-006-9138-z Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants of social behavior. Personality and social psychology review, 8 (3), 220•247. doi:10.1207/s15327957pspr0803_1 Tolin, D. F., Lohr, J. M., Lee, T. C., & Sawchuk, C. N. (1999). Visual avoidance in specific phobia. Behaviour research and therapy, 37(1), 63-70. doi:10.1016/S0005- 7967(98)00111-9 Weinstein, N., Ryan, W. S., DeHaan, C. R., Przybylski, A. K., Legate, N., & Ryan, R. M. (2012). Parental autonomy support and discrepancies between implicit and explicit sexual identities: Dynamics of self•acceptance and defense. Journal of personality and social psychology, 102 (4), 815•832. doi:10.1037/a0026854 Yuen, K., & Ickes, W. (1984). “Prudes” and “pornographiles”: Effects of subject and audience attitudes on the viewing and rating of pornographic materials. Journal of Social and Clinical Psychology, 2(3), 215-229. doi:10.1521/jscp.1984.2.3.215

27

6. Appendices

Appendix A – The Modern Homonegativity Scale (Morrison & Morrison, 2002) and its French translation

1. Many gay men use their sexual orientation so that they can obtain special privileges. De nombreux hommes homosexuels utilisent leur orientation sexuelle pour obtenir des avantages. 2. Gay men seem to focus on the ways in which they differ from heterosexuals, and ignore the ways in which they are the same. Les hommes homosexuels semblent se concentrer sur les manières de faire qui les différencient des hétérosexuels. En revanche, ils ne tiennent pas compte des manières de faire identiques à celles des hétérosexuels. 3. Gay men do not have all the rights they need.* Les hommes homosexuels n’ont pas tous les droits dont ils ont besoin. 4. The notion of universities providing students with undergraduate degrees in Gay and Lesbian Studies is ridiculous. L’idée que les universités fournissent des diplômes en lien avec l'étude des gays et lesbiennes est ridicule. 5. Celebrations such as “ Day” are ridiculous because they assume that an individual’s sexual orientation should constitute a source of pride. Les fêtes telles que la "Gay Pride" sont ridicules parce qu’elles partent du principe que l’orientation sexuelle d’un individu devrait constituer une source de fierté. 6. Gay men still need to protest for equal rights.* Les hommes homosexuels ont encore besoin de manifester en faveur de l’égalité des droits. 7. Gay men should stop shoving their lifestyle down other people’s throats. Les hommes homosexuels devraient arrêter d’imposer leur mode de vie aux autres. 8. If gay men want to be treated like everyone else, then they need to stop making such a fuss about their sexuality/culture. Si les hommes homosexuels veulent être traités comme tout le monde, alors il faut qu’ils arrêtent de faire toute une histoire de leur sexualité/culture. 9. Gay men who are “out of the closet” should be admired for their courage.*

28

Les hommes homosexuels qui sont “sortis du placard” devraient être admirés pour leur courage. 10. Gay men should stop complaining about the way they are treated in society, and simply get on with their lives. Les hommes homosexuels devraient arrêter de se plaindre sur la façon dont ils sont traités dans la société et devraient simplement continuer à vivre leur vie. 11. In today’s tough economic times, Canadians’ tax dollars shouldn’t be used to support gay men’s organizations. Dans cette période économique difficile, l’argent des contribuables Suisses ne devrait pas être utilisé pour soutenir des organisations en faveur des hommes homosexuels. 12. Gay men have become far too confrontational in their demand for equal rights. Les hommes homosexuels sont devenus beaucoup trop agressifs dans leur demande pour l’égalité des droits. ______Note: * items are reverse-scored

29

Appendix B – Sample photographs of male homosexual and heterosexual couples B.1 Heterosexual couples

30

B2. Homosexual couples

31

Appendix C – Scatterplot of homonegativity scores as a function of IAH