<<

UNDERSTANDING CONTRAPOWER IN

A thesis presented to

the faculty of

the College of and Sciences of Ohio University

In partial fulfillment

of the requirements for the degree

Master of Science

Melissa M. Smart

November 2007 2

This thesis titled

UNDERSTANDING CONTRAPOWER IN SEXUAL HARASSMENT

by

MELISSA M. SMART

has been approved for

the Department of Psychology

and the College of Arts and Sciences by

______

Paula M. Popovich

Associate Professor of Psychology

______

Benjamin M. Ogles

Dean, College of Arts and Sciences 3 Abstract SMART, MELISSA M., M.S., November 2007, Industrial/Organizational Psychology

UNDERSTANDING CONTRAPOWER IN SEXUAL HARASSMENT (130 pp.)

Director of Thesis: Paula M. Popovich

Sexual harassment (SH) in organizations is a complex problem that has been of interest for nearly 30 years. A number of theories have been proposed to explain SH, and

while none of them have had extensive support, power is a consistent element. Both

traditional and non-traditional forms of SH exist, including contrapower harassment in

which the harasser appears to hold less formal power than the victim. The purpose of this

study was to investigate of contrapower SH in order to gain a more thorough

understanding of SH. Participants read one of 8 vignettes, in which several characteristics

had been varied. Participants answered questions regarding whether harassment had

occurred, and if so, who in the interaction was at fault. Results revealed evidence for both

traditional and contrapower SH. However, no significant findings for participant gender

or exposure to training/education were found.

Approved: ______

Paula M. Popovich

Associate Professor of Psychology

4

Table of Contents

Page

Abstract...... 3 List of Tables ...... 6 List of Figures...... 7 Understanding Contrapower in Sexual Harassment ...... 8 Sexual Harassment Background ...... 9 Legal Definitions of Sexual Harassment ...... 10 Perceptions of Sexual Harassment...... 14 Legal Aspects of Perceptions of Sexual Harassment...... 14 Differences in Perceptions of Sexual Harassment ...... 15 Summary ...... 18 Models of Sexual Harassment ...... 19 Biological/Evolutionary Models ...... 19 Sociological Models...... 21 Organizational Models ...... 22 Individual Models ...... 25 Feminist Models...... 26 Summary ...... 27 The Influence of Power in Sexual Harassment...... 28 Summary ...... 34 Traditional and Non-traditional Sexual Harassment...... 35 Traditional Sexual Harassment ...... 35 Contrapower Sexual Harassment ...... 37 Summary ...... 42 Organizational vs. Academic Settings ...... 42 Summary ...... 46 The Present Study ...... 46 Hypotheses...... 48 Method...... 50 Participants...... 50 Design ...... 50 Materials ...... 51 Procedure ...... 53 Results...... 53 Additional Analyses...... 77 Summary ...... 85 Discussion...... 85 General Findings...... 85 Position Power ...... 86 Gender...... 91 Harasser Gender...... 91 Participant Gender ...... 93 5 Additional Findings ...... 94 Exposure to Sexual Harassment Training ...... 94 “No Harassment” Conditions ...... 94 Limitations and Future Research ...... 95 Implications...... 97 Conclusions...... 98 References...... 100 Appendix A...... 105 Appendix B...... 107 Appendix C...... 109 Appendix D...... 111 Appendix E ...... 113 Appendix F ...... 118 Appendix G...... 120 Appendix H...... 123 Appendix I ...... 124 Appendix J ...... 126 Appendix K...... 128 6 List of Tables

Page

Table 1 Multivariate Analysis of Variance and Means for Type of Vignette...... 54

Table 2 Multivariate Analysis of Variance and Means for Type of Vignette,

Control Vignettes Only...... 55

Table 3 Multivariate Analysis of Variance and Means for Type of Harasser...... 56

Table 4 Multivariate Analysis of Variance for Professor vs. Student Harassers...... 59

Table 5 Multivariate Analysis of Variance for Harassment Vignettes Only...... 62

Table 6 Multivariate Analysis of Variance for Male vs. Female Harassers...... 65

Table 7 Multivariate Analysis of Variance for Harasser Gender x Position

Power Interaction...... 69

Table 8 Multivariate Analysis of Variance and Means for Participant

Training Experience...... 70

Table 9 Multivariate Analysis of Variance for Participant Gender...... 71

Table 10 Categorizations of Open-Ended Responses...... 74

Table 11 Multivariate Analysis of Variance and Means for Type of Initiator,

“No Harassment” Only...... 76

Table 12 Multivariate Analysis of Variance for Professor vs. Student

Initiators, “No Harassment” Only...... 77

Table 13 Multivariate Analysis of Variance for Male vs. Female Initiators,

“No Harassment” Only...... 79

Table 14 Multivariate Analysis of Variance for Initiator Gender x Power

Interaction, “No Harassment” Only...... 82 7 List of Figures Page

Figure 1 Vignette Conditions Included in the Current Study...... 51 8

Understanding Contrapower in Sexual Harassment

Sexual harassment is a prevalent problem in organizations, with 44% of women and 19% of men reporting such harassment (U.S. Merit Systems Protection Board, 1995).

In order to deal with this problem, it is important to first understand the issue, including the major characteristics and causes of sexual harassment. Sexual harassment can take many forms, both traditional and non-traditional, but power is a recurring theme in most descriptions of sexual harassment, creating a focal point for the current research.

Wilson and Thompson (2001) state that when examining power and sexual harassment, traditionally men are the harassers of women, and these men are often in positions of formal power over their victims. Quite often, sexual harassment is prevalent where there are large formal power differentials (Cleveland & Kerst, 1993). Formal power usually refers to situations in which an individual occupies (by appointment or election) a position that allows him/her to have control over others. For example, a CEO has formal power over his or her secretary. Formal power refers to power within an organizational context. Cleveland and Kerst, in their discussion of societal, organizational, and individual powers, also concluded that men typically hold more power than do women, which contributes to the higher rates of female victimization.

In addition to the more traditional forms of sexual harassment (e.g. a man harassing a woman or someone with power harassing someone with less power), there are also non-traditional forms of sexual harassment, including the harassment of men, same- sex harassment, and contrapower harassment. Contrapower sexual harassment arises in situations where the harasser has less formal power than the victim (Benson, 1984), and 9 is an excellent example of how other sources of power may contribute to an incident of sexual harassment.

Power in sexual harassment can stem from several different sources including biological or evolutionary (Tangri & Hayes, 1997), sociological/societal (Wilson &

Thompson, 2001), organizational, and individual sources (Ragins & Sundstrom, 1989;

Cleveland & Kerst, 1993). Power is an important aspect of sexual harassment behavior and incidents. Perhaps a more comprehensive understanding of power, including non- traditional power relationships (e.g., contrapower), can be applied to sexual harassment to help us to better understand the various forms sexual harassment can take. Examining the particular characteristics (e.g., setting, power differences, gender of victim/harasser) of organizations that give rise to various forms of sexual harassment may also allow us to predict when there is a likelihood that these behaviors may become a problem.

Specifically, the unique form of sexual harassment, contrapower sexual harassment, was investigated in the current study. can interpret the Equal

Employment Opportunity Commission (EEOC) guidelines on sexual harassment in different ways. Furthermore, individuals have differing familiarity with the various forms of sexual harassment (i.e., traditional versus non-traditional). Because of these two factors, contrapower harassment is particularly interesting because it may or may not be considered harassment, either legally or perceptually.

Sexual Harassment Background Sexual harassment has only come into the public eye and into research within the last 30 years or so. The first survey by the U.S. Merit Systems Protection Board conducted in 1980 showed that sexual harassment in the workplace was perceived to be a 10 problem (USMSPB, 1995). According to the 1995 U.S. Merit Systems Protection

Board, over 40% of women and nearly 20% of men reported experiencing some form of sexually harassing behaviors. The report states that because these numbers have not significantly changed since an earlier administration of the survey (1987), sexual harassment is still an issue that needs to be addressed. While sexual harassment has undoubtedly existed prior to the 1980s, it was in this decade that research investigating this phenomenon began to become more common (e.g., Reilly, Carpenter, Dull, &

Bartlett, 1982; Popovich, Licata, Nokovich, Martelli, & Zoloty, 1986; Abbey, 1987;

Pryor, 1987).

With the establishment of the Equal Employment Opportunity Commission guidelines on sexual harassment, organizations were able to adopt a common legally- based definition of sexual harassment to use when dealing with internal complaints. The

EEOC guidelines took sexual harassment from an unmentioned part of every day life for some employees to a punishable offense.

Legal definitions of sexual harassment There are two general forms of sexual harassment: quid pro quo and hostile environment (EEOC, 1980). Quid pro quo sexual harassment involves an economic impact on the victim. Specifically, “unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature… when (1) submission to such conduct is made either explicitly or implicitly a term or condition of employment,

(2) submission to or rejection of such conduct by an individual is used as the basis for employment decisions affecting such individual, or (3) such conduct has the purpose or effect of unreasonably interfering with an individual’s work performance or creating an 11 intimidating, hostile, or offensive working environment” (pp. 74677). The EEOC guidelines on sexual harassment can be found in Appendix A.

The third descriptor in the EEOC guidelines on sexual harassment refers to hostile environment sexual harassment in which there is no direct economic impact on the victim. Examples of quid pro quo sexual harassment include behaviors such as a woman being threatened with termination unless she sleeps with her boss, or making a sexual act a term of being hired. Examples of hostile environment sexual harassment include sexual jokes, offensive pictures, and repeated requests for sexual favors which do not affect employment status. Quid pro quo sexual harassment is often perceived to be more serious than hostile environment sexual harassment, but both types of behavior are actionable under the law. Even with these legal guidelines, sexual harassment is not always clear because of its perceptual nature.

An important, yet controversial, component of sexual harassment is intention:

That is, did the harasser intend to harass the victim, or was there simply a misunderstanding. It is important to note that legally, intent on the part of the harasser does not matter, whereas perceptually, intent does matter. Specifically, on the one hand, when examining the totality of the circumstances regarding an allegation of sexual harassment, whether or not the harasser intended to harass the victim does not matter. On the other hand, when an individual perceives that he or she has been harassed, it is likely that the victim considers the perceived intent of the harasser (MacKinnon, 1979;

USMSPB, 1995).

Another legal aspect of sexual harassment involves who within an organization may be considered a victim or a harasser. Harassers can be men and women, as can their 12 victims. Also, the victim need not be the opposite sex of the harasser (i.e., same-sex harassment also qualifies as sexual harassment). Furthermore, a victim of sexual harassment does not necessarily have to have been involved in the incident itself (third party harassment). Third party harassment means that any individual who is adversely impacted by the sexually harassing behaviors, even if he or she is not the intended target, can be considered a victim. Finally, the harasser can have a variety of relationships to the victim: The harasser may be the victim’s supervisor, co-worker, subordinate, an agent of the employer (e.g. a supplier), or even an individual who is not employed by the organization. When investigating allegations of sexual harassment, all of the circumstances surrounding the incident are considered on a case-by-case basis (EEOC,

2001).

According to the U.S. Merit Systems Protection Board (1995) report, few victims of sexual harassment take formal action against their harassers. The report states that although 78% of the survey respondents were aware of the procedure for taking formal action against a harasser (the same was also true of 76% of harassment victims), only 6% of the survey respondents reported having taken formal action. The figures reported by the EEOC show that only 3% of the complaints filed with the commission were sexual harassment-related discrimination complaints, further demonstrating the limited number of cases that receive formal action. The types of formal action taken by the respondents to the survey of federal employees fall into several categories: 42% requested an investigation by the organization, 30% filed a discrimination complaint or suit, 25% filed a grievance or adverse action appeal, 14% requested an investigation by an organization outside of the employer, and 17% took some other form of action. Some individuals took 13 more than one formal action. Of those who chose not to take any formal action in response to harassment, the USMSPB report states that it was a very low percentage who chose not to take action because either they felt the supervisors would not be supportive

(6%) or because they did not know how (5%). The report notes that these figures are important because they indicate that employees largely feel that management will be supportive and are aware of the channels for reporting harassment.

When allegations of sexual harassment are brought into the legal arena, the

“reasonable person” standard is generally employed for determining whether harassment has occurred. This standard means that the deciding parties (e.g., a judge or a jury) should consider what a “reasonable person” would do in a situation with similar circumstances

(Gutek, O’Connor, Melancon, Stockdale, Geer, & Done, 1999). Gutek, et al. describe the origins and evolution of the reasonable person standard .The authors note that the reasonable person standard began with the reasonable man standard which was employed in the 1700s and was intended to refer to all people, not only men. The reasonable man standard evolved into the reasonable person standard. However, the meaning remained the same; specifically, that it is “gender-free”.

Gutek, et al. detail (1999) several criticisms of the reasonable person standard when applied to sexual harassment and how, out of the criticisms, the reasonable woman standard arose. In an attempt to make decision-makers focus more explicitly on the perspective of a female victim of harassment, the reasonable woman standard states that individuals should consider what a reasonable woman would do in a situation similar to that of the situation being considered. The purpose of this focus is to account for circumstances which men may consider harmless, but women consider harassing. The 14 reasonable person and reasonable woman standards are both legal standards used when determining whether sexual harassment has occurred. However, the “reasonableness” of these standards is very much perceptual in nature. What one individual considers reasonable may differ from what another individual considers reasonable.

It is important to note that regardless of the various standards employed by the court system, each allegation of sexual harassment is evaluated on its own characteristics, such that no one ruling will apply across instances. Each incident is considered on a case- by-case basis, which ultimately involves the perceptual nature of sexual harassment

(EEOC, 1980).

Perceptions of sexual harassment Legal aspects of perceptions of sexual harassment. When deciding whether an incident qualifies as sexual harassment, the perceptions of the individuals involved must be taken into account. Depending on the form of sexual harassment, and who is judging the incident, different perspectives will be acknowledged. In the legal arena, the victim’s of the incident receives greater weight. Specifically, regardless of whether the harasser intended for the behaviors to be sexually harassing, if the victim feels as though he/she has been harassed, the fact that the individual feels that harassment has occurred is enough to warrant investigation (Popovich, 1988). In quid pro quo sexual harassment, the harasser’s intention is fairly straight-forward. When examining hostile environment harassment claims, however, intentions are less obvious. In these cases, the perceptions of both parties involved must be considered. For example, if an individual who is telling sexually explicit jokes in the workplace does not have the intention of sexually harassing or offending those around him, but a group of female co-workers feel as though they have 15 been harassed, the appropriate disciplinary action that is needed (if any) is less clear and the perceptions of all the individuals involved must be taken into account.

Individual differences in perceptions of sexual harassment. A major example of how perceptions can affect sexual harassment outcomes involves gender differences, such that men and women often differ on their perceptions of what constitutes sexual harassment (Blumenthal, 1998). In his meta-analysis, Blumenthal reported that while there were many gender effects found across his sample of studies, a number of characteristics impacted the size of gender effects, including; the type of protocol (e.g., narratives), type of stimulus materials (e.g., legal setting), time of study (recent or not-so- recent), and type of participants (students or workers).

Rotundo, Nguyen, and Sackett (2001) also conducted a meta-analysis on gender differences in perceptions of sexual harassment. Unlike Blumenthal (1998), these researchers did not combine all forms of sexual harassment. Rotundo and colleagues gathered information on the types of behaviors that previous researchers have been using to describe instances of sexual harassment and broke them into 7 categories. Six of these categories (derogatory attitudes- impersonal, derogatory attitudes- personal, unwanted pressure, sexual propositions, physical sexual contact, and physical nonsexual contact) corresponded to hostile environment sexual harassment, while the final category

(sexual coercion) corresponded to quid pro quo sexual harassment. Their results showed that while women tended to perceive more behaviors as sexual harassment, the gender differences seemed to occur more often with the less severe categories, such as physical nonsexual contact and sexual propositions. Gender differences were much smaller for the sexual coercion harassment behaviors. 16 When taken together, the results of these two meta-analyses indicate that, while there are gender differences in perceptions of sexual harassment, these differences are variable and may not be as severe as they were once believed to be.

Gender differences in perceptions of sexual harassment have changed over time.

The U.S. Merit Systems Protection Board report (1995) includes a table indicating the percentages of men and women who endorsed six general categories of behaviors as constituting sexual harassment. The categories included; pressure for sexual favors, deliberate touching and cornering, suggestive letters, calls, or other materials, pressure for dates, suggestive looks or gestures, and sexual teasing, jokes, or other remarks. The report presents figures from the 1980, 1987, and 1994 surveys.

In general, a greater percentage of men and women endorsed pressure for dates, suggestive looks or gestures, and sexual teasing, jokes, or other remarks as constituting sexual harassment in the 1987 and 1994 surveys than did in the 1980 survey. The increase was often nearly 20% more respondents endorsing these behaviors as harassment. Furthermore, while the male respondents had fewer endorsements across all categories and years, the figures from 1987 and 1994 are much closer to the figures reported by female respondents to the surveys.

Overall, both male and female respondents to the surveys report more behaviors as constituting sexual harassment in the later surveys. Also, while male respondents reported more the behaviors as harassment, these rates are still less than those of the female respondents. There appears to be the greatest disparity between male and female endorsements for the less serious categories, such as suggestive looks or gestures, and sexual teasing, jokes, or other remarks. It has been suggested that sexual harassment 17 behavior reflects a more general social phenomenon, that of men’s misperceptions of women’s friendliness as sexual interest (Stockdale, 1993; Abbey, 1987).

Stockdale (1993) has discussed ways in which women’s friendliness can be misperceived by some men as what the author labels as “sexiness.” This misperception by men may be misconstrued as either an invitation for sexual behavior or at the very least, an acceptance of such behavior, which opens the door for sexual harassment. Some women have very outgoing and friendly personalities, which lead them to engage men in conversation in an open and forward manner. From the perspective of some men, the women pursuing pleasant conversation with them can be seen as exhibiting “sexiness” rather than mere friendliness. Depending on how the woman responds to the man’s advances, it is not difficult to see how harmless could escalate to perceived sexual harassment. Stockdale stated that it could also come down to characteristics of the individual man as to whether the woman will perceive flirtatious behavior as sexual harassment. Specifically, if a man has expressed degrading or controlling views about women, his behavior may be perceived as harassment. However, if a man has expressed generally positive opinions about women, his behavior may be more likely to be viewed as harmless flirtation. Throughout her work, Stockdale cites research by Abbey also dealing with misperceptions of behavior.

Abbey (1987) conducted two studies to examine participants’ past experiences with misperceptions of friendly behavior as sexual interest. The results of the two studies showed that, while misperceptions were common, they were most often of the type that men were misperceiving women’s friendly behavior as sexual interest. Male and female participants reported differing results from the misperceptions. When discussing how the 18 participants became aware of the misperceptions, Abby stated that women were more often physically touched by their misperceivers (thus realizing that the men had misunderstood the women’s intentions), whereas men were more often made aware by others or by the women themselves of the misperception. In response to the misperception, women were likely to stop the touching, although men were more likely to either go along with the behavior or simply ignore it.

It is clear that perceptions of sexual harassment are important to consider when evaluating incidents of sexual harassment. One aspect of such incidents that can become important is the relative perceptions of power that each individual has. For example, an individual who is being subjected to potentially harassing behaviors by someone who they perceive to have a great amount of power over them may have a drastically different reaction to the behavior than someone who is being subjected to the same behaviors by an individual with little or no power over him or her.

Summary. While gender plays an important role in sexual harassment, the relative power that individuals possess can have a substantial impact on perceptions of potentially harassing behaviors. Popovich, Licata, Nokovich, Martelli, and Zoloty (1986) assessed the incidence and perceptions of sexual harassment behaviors in a way that begins to hint at the influences of power on perceptions of sexual harassment. In the second of the two included experiments, participants were more likely to state that a behavior was harassment if the perpetrator was in a supervisory position over the victim. Others (e.g.

Cleveland & Kerst, 1993) have also noted the importance of formal, position power.

Specifically, that while co-worker harassment is more common, supervisor harasser is deemed more severe. While position power in an organization is the most salient, there 19 are several other models of sexual harassment that must be considered when examining sexual harassment and power. Each of these models can be examined strictly as a of sexual harassment or as a model of power. Combining the sexual harassment and power aspects of these models will help us to investigate this phenomenon.

Models of sexual harassment There are five major models of sexual harassment which will be reviewed: biological/evolutionary (Tangri & Hayes, 1997), sociological (Wilson & Thompson,

2001), organizational (Cleveland & Kerst, 1993), individual (Tangri & Hayes, 1997), and feminist (Lips, 1991). While none of these models are specifically being endorsed in the current examination of contrapower sexual harassment, these models represent some of the theories that have been advanced to explain why sexual harassment occurs. Not all of these models have received empirical support, but some have important implications for understanding why harassment continues to exist in today’s organizations.

Biological/Evolutionary models. The biological or evolutionary model relies on sex and aggression components, as these are two of human beings’ primary drives

(Tangri & Hayes, 1997, p.113). This model describes sexual harassment as coming from a biological basis. Specifically, since men have stronger sex (hormonal) drives, the resulting behavior is the sexual harassment of women. Tangri and Hayes describe two of the prominent biological models of sexual harassment. The hormonal model states that sexual harassment is merely a reflection of human hormonal drives. That is, men become harassers because they are simply following their hormonal desires to act on their urges.

If the hormonal model reflected reality, however, older women at their sexual peak would 20 also be likely harassers and this does not appear to be the case, according to the authors. There is little empirical support for the hormonal model.

An additional variation of the biological model is the Evolutionary Adaptation model. Tangri and Hayes (1997) state that the sexual harassment of women under this model occurs because men strive to impregnate as many women as possible to further their . It is beneficial for men to make sexual advances toward as many women as possible. The authors write that women must be more selective with regards to sexual encounters, due to the investment required to successfully raise children. Because of this investment, women view repeated sexual advances in a more negative light than do men.

The Evolutionary Adaptation model creates a conflict of interest between men and women, and does not explain all observed sexual harassment behaviors or even other logically predictable behaviors. Like the hormonal model, this model has not been well supported (Tangri & Hayes, 1997). The authors conclude with a statement summing up the evidence for these models: “At best, we can only conclude that the evidence for the natural/biological model, with its contradictory derivations, is mixed and, worse, not falsifiable” (pp. 116).

Finally, although not specifically tied in to either the hormonal model or the

Evolutionary Adaptation model, another biological aspect of sexual harassment and power includes physical size. For instance, because men are generally physically larger than women, it is fairly simple to imagine how a woman could feel threatened into submitting to sexually harassing behaviors from a man because she is intimidated by his physical size relative to her own size. 21 Sociological models. The second major model is the sociological approach to sexual harassment. Wilson and Thompson (2001) describe Lukes’ three-dimensional, sociological model of power and apply it to sexual harassment. The first dimension is simply one person’s interests prevailing over those of another individual in decision making. Wilson and Thompson incorporate this dimension into sexual harassment such that men use their societal power to give weight to their sexual advances, winning out over the protesting interests of women.

Lukes’ second dimension adds a structural element to the mix, such that powerful individuals have the ability to prevent action being taken on issues for which they feel are unimportant or pushing action on subjectively important issues. For example, upper level management often has the power to make decisions on whether disciplinary action will be taken against harassers within an organization. If managers, who may also be harassers, do not want to address the problem within the organization, it is likely that sexual harassment will never be dealt with.

The third, and final, dimension that Wilson and Thompson discuss moves to the

“deep structures” of power. This is the idea that the structure of power is so ingrained that individuals without power do not question the system; therefore, sexual harassment is not a reported issue. For example, the dominance of men over women is so ingrained in the organization that when a man asserts his power over a woman in a harassing way, it seems natural and she does not take his behavior as a form of sexual harassment: She has learned to cope with this type of behavior as an every-day part of her work life.

Wilson and Thompson’s explanation and utilization of Lukes’ three-dimensional model of power illustrate how power roles can become institutionalized within . 22 Furthermore, the model illustrates how women may become desensitized to sexual harassment stemming from power and may not report harassing behaviors. Wilson and

Thompson are not the only authors to suggest that the status structure can become ingrained in society. Tangri and Hayes (1997) state that, “There is general agreement in the literature about the characteristics of the sex stratification system and the socialization patterns that maintain it” (pp. 120). The main idea behind the sociological model has received support in the literature.

Organizational models. The third model of sexual harassment is the organizational model. Much research as been conducted examining the influence of position power on sexual harassment. According to Cleveland and Kerst (1993), organizational power refers to the power structure within an organization; specifically, looking at who holds which positions of power. Ragins and Sundstrom (1989), in their discussion of gender and power in organizations, discuss aspects of an organization that can influence who rises to positions of power. These aspects include entry into the organization and promotion. The manner used to publicize job openings to attract applicants can affect the types of people who will be drawn to the organization. One example of this process is the “old boy” network, where only certain people are made aware of important information. Promotion methods within an organization also clearly influence the individuals with power.

Both Cleveland and Kerst (1993) and Ragins and Sundstrom (1989) focus on the idea that through various operations in an organization, men tend to hold more positions of power within organizations. Individuals with power may feel that they have the authority to make demands on less powerful individuals, and less powerful individuals 23 may then feel pressure to give in to such demands (Eagly, 1983). When applied to sexual harassment, the organizational model demonstrates how men can become sexual harassers of women. Men move higher and more quickly through the organizational hierarchy, which affords them more power to make sexual demands of women who will likely be in a greater number of subordinate positions. These women may feel compelled to comply with requests for sexual behaviors for fear of repercussions, including losing their jobs.

Another component of the organization that can lead to sexual harassment behaviors involves the proportion of each gender employed by the organization. Gender proportion becomes important in the workplace when examining phenomena such as the sex-role spillover effect. Gutek and Cohen (1987) described sex-role spillover as an individual’s work life being affected by his or her sex-roles. They write that individuals in predominately opposite-sex typed occupations are generally the most affected by sex- role spillover. Men and women have drastically different sex roles, with women being described as warm, friendly, homemakers and men being described as strong, professional, assertive, breadwinners. When these characteristics meet the workplace, expectations for behavior arise. For example, people are more adept at interacting with people in ways that are congruent with their assumed sex-roles. Problems develop when the expected sex roles do not apply to someone and others do not adapt. Sex-role spillover implies that a woman working in a predominately male area, management or otherwise, will often be seen as a woman first, and a worker second. Gutek draws from work by Kanter in describing the importance of the gender breakdown in the workplace. 24 Kanter (1977) describes 4 types of groups of people that can be found in organizations based on the ratio of men to women. Uniform groups consist of entirely one sex. Skewed groups have a few members of one sex and many members of the other.

Tilted groups move closer to equality but are not quite to that point. In this type of group, one group becomes the majority, and the other the minority. Finally, in a balanced group, the sexes are represented in equal numbers. The group of interest in sex-roll spillover cases is the skewed group. Kanter writes that the few “token” members in the group will not be seen as individuals, but as representatives, or symbols, of their group (pp. 208).

When applied to sex-roles, the few women in male-dominated occupations (e.g., management) become symbols of womanhood, and not of workers.

Gutek and Cohen (1987) discuss the effects of sex-role spillover as being more negative for women than for men, such that women experiencing sex-role spillover are more likely to be the targets of sexual advances and sexual harassment by men. Men in typically female-typed jobs are much less likely to be the targets of sexual behaviors from women. This finding is partially because the effects are found for individuals who work in opposite-sex typed jobs, and men are not usually the minority, especially if management is being studied. Women who are in predominately male jobs, however, are very visible as being a minority.

Sex-role spillover is a prominent theory in sexual harassment research. The sexual harassment of women in organizations is theorized to develop because men fail to see past a female employee’s sex, viewing her as little more than a possible target for his sexual interest. Instead of perceiving his female co-workers as equals and professionals, a 25 man sees female co-workers as sex objects. This misperception of women’s roles in the workplace opens the door to sexual harassment, regardless of whether it was intentional.

Tangri and Hayes (1997) note that the principles behind sex-role spillover have been supported in the research literature. The following predictions of the theory have received support: Women working in male-typed jobs perceive that they are treated differently, and that this behavior is directed at their sex and is perceived as harassment; women working in female-typed jobs are less likely to report discriminatory or harassing behavior, as are women who work in balanced environments.

Individual models. The fourth model of sexual harassment is the individual model. This model deals with which characteristics will make someone more or less likely to become a harasser or a victim. Tangri and Hayes (1997) state that individual differences are generally examined from two different points of view. The first approach studies the differences between harassers and non-harassers, and the second approach investigates the characteristics of individuals who perceive incidents of sexual harassment and what makes these individuals decide whether the incident truly was harassment. Pryor (1987) developed the Likelihood to Sexually Harass scale to examine the characteristics of harassers. The scale was developed for men, as they are primarily the harassers. Results showed that the scale was correlated with other scales and measures including; likelihood to , sex-role stereotyping, and acceptance of interpersonal violence. Harassers also seemed to lack the ability to view incidents from the perspective of others. This research demonstrates that there do appear to be individual difference variables that distinguish between harassers and non-harassers. 26 Tangri and Hayes (1997) also address the characteristics of individuals and incidents that make it more or less likely for behavior to be considered sexual harassment. This discussion again centers on the previously discussed gender differences in perceptions of potentially sexually harassing behaviors. Further evidence for this gender difference in perceptions of behaviors can be found in Levesque, Nave, and Lowe

(2006). In their study investigating gender differences in inferring sexual interest, the researchers found that after two individuals would have a brief conversation and rate the interaction, men would tend to oversexualize the interactions to a greater extent than would women. The authors took this result as evidence to support the general finding that men think in a more sexual manner than do women. Similar to the previous discussion of gender differences in perceptions of sexual harassment, the results of this study demonstrate how women’s merely friendly behaviors can be interpreted as sexual interest by men, further providing evidence for gender differences in perceptions. This miscommunication may lead men to behave in a manner that is perceived to be sexual harassment by women.

Feminist models. The fifth, and final, model is the feminist approach. Lips (1991) states that a basic assumption underlying the feminist model is that men and women hold different amounts of power. This concept is similar to that proposed in the sociological and organizational models of power in which the structure of power is deeply ingrained.

The feminist approach also asserts that the power structure is deeply ingrained in society such that men have more power. Feminists work to improve the power status of women in society. Lips describes the debate among feminist theorists as to what the appropriate move is regarding the power differential between the genders. This debate divides power 27 into two types. The first is “power over” others, or the “bad” power. The second is the

“power to” accomplish one’s goals, or the “good” power. Lips asserts that feminists aim to achieve the good power, while not necessarily achieving the bad power. The author writes that this goal will bring each individual the power to control his or her life, while not placing any individual over another. These two types of power cannot be completely separated, however, because as Lips states, once an individual grows in the ability to achieve personal goals and begins feeling more empowered, there is an increased likelihood that this individual will challenge authority and behave in ways that can mirror the “bad” power over others. The feminist model acknowledges that while the power to achieve goals is the desirable power, power over others is necessary to some extent to avoid a structureless society.

The power inequality between the genders can lead to sexual harassment. Lips

(1991) describes a situation in which a subordinate female must participate in sexual acts with her boss to avoid being fired. The man has so much power over this woman that she stays silent about the incident and begins to feel as though this man has made her powerless through his actions. This incident demonstrates the idea that men can continue to hold women in powerless positions which can cause them to submit to sexual advances.

Summary. Although these models seem to point to the conclusion that men, and generally only men, are harassers, it is important to note that not all men are, or will become, sexual harassers. The previously-mentioned Likelihood to Sexually Harass scale developed by Pryor (1987) can be used to determine which men possess characteristics that may put them at greater risk for sexually harassing others. The idea for this research 28 stemmed, in part, from literature examining responses to a scale addressing the likelihood that a man would rape a woman if there was no chance that he would be caught and punished. Researchers in this area noted that extreme forms of sexual harassment can be similar to rape, suggesting that perhaps sexual harassers and rapists may also be similar in some respects. In order to investigate this relationship, Pryor began working on the development of the LSH scale to determine characteristics of sexual harassers. Like the Likelihood to Rape scale, the LSH scale included a series of scenarios followed by a group of possible actions that could be taken. One of the possibilities involves the man sexually harassing a woman without the possibility of negative consequences. In a series of 3 studies, Pryor found that men who score highly on the LSH scale are likely to hold adversarial sexual beliefs, have difficulty empathizing with others, have higher proclivities to rape, be higher in authoritarianism and lower on

Machiavellianism, hold negative feelings about sexuality, describe themselves in ways that differentiate themselves from , and act in a sexualized manner towards others when the situation provides an opportunity or excuse for the behavior. The LSH scale provides a way to determine which men may be likely to become sexual harassers.

By examining the list of characteristics provided above, it is clear that not all men will qualify as harassers.

The influence of power in sexual harassment The five models described above offer a wide variety of ways that sexual harassment can be conceptualized and described. One common thread that can be drawn through each of the models is that power can play a central role in sexual harassment. In the biological/evolutionary model of sexual harassment, power can manifest itself in the 29 physical size advantage that men generally have over women, since men tend to be both larger and stronger. In the sociological model, power is seen as a masculine trait that is a socially constructed part of male sex roles. Therefore, in this model, too, men have more power than women. The organizational model and the individual model both look at characteristics of those who hold power, either (more formally) as a part of an organization or (less formally) as an individual.

One model that has been widely used in the literature that includes both formal and informal power is French and Raven’s (1959) model of individual sources of power.

This model has been used to explain power at various levels, especially the organizational and individual levels. The sources of power included in this model are legitimate, reward, coercive, referent, expert, and informational power. Each of these sources of power can be tied to the various models of sexual harassment at different levels. For example, legitimate, reward, and coercive powers can all fit within the organizational model, such that individuals higher up in the organization’s hierarchy of power will, by virtue of their positions, possess legitimate power. An increase in formal power comes with more opportunities to use reward and coercive powers.

A more thorough description of each of French and Raven’s (1959) and Raven’s

(1965) sources of power will be helpful in demonstrating how each of these powers fits into the four classifications.

Legitimate power, or position power, is held by individuals with a formal, hierarchical position of power and can be specifically applied to the organizational models of sexual harassment. For instance, a professor has legitimate power over his or her students because the students recognize that their professors are in a position of 30 authority with influence over grades and other course outcomes (Dziech & Weiner,

1984). Operating under the organizational model in which a harasser has position

(legitimate) power over his/her victim, a professor might make advances toward a student while at school, as well as make suggestive phone calls to the student. Reward power stems from the ability to bestow rewards, and similarly, coercive power comes from the ability to punish others. An example of the use of reward power in the organizational model would be that a superior tells his employee that if she has sex with him, he will promote her to a higher-paying position. A simple example of coercive power in an organization is the situation of a manager threatening the job of his secretary unless she agrees to his demands. Legitimate, reward, and coercive power are all considered positional powers: supervisors are generally seen as having these sources of power from the perspective of a subordinate employee.

Referent power is the ability to make others want to be like you. This type of power occurs when an individual identifies with someone else to the extent that he or she strives to be like that person, thus bestowing referent power upon him or her. Out of all of the models of sexual harassment, referent power fits most easily within the individual model. For example, an individual possessing referent power over another would be able to use the fact that the victim wants to identify with him/her to extract sexual favors.

While the victim chooses to comply with demands (because he or she wants to be like the harasser), the harasser could easily use this power to encourage participation in harassing behaviors.

Expert power is based on the acquisition of skills or knowledge. An individual has this power if he or she can demonstrate more knowledge than others. A professor holds 31 this form of power over students, in addition to legitimate power. A supervisor could deny an employee access to necessary information unless the employee submits to sexual behaviors. This situation is an example of how this individual power (thus applying to the individual models of sexual harassment) can be used to perpetuate sexual harassment.

The final type of individual power is informational power, or the control over communication and persuasion. One can hold a great deal of power if he or she has the ability to use communication, information, and persuasive arguments to his or her advantage. While expert and informational powers may seem as though they would be less likely to be abused in sexual harassment, these powers could be used as bargaining chips by a harasser in his or her attempt to gain sexual favors.

Referent, expert, and informational power all tend to be individual powers and can, therefore, be applied to the individual model of sexual harassment. Each of these sources of power can be viewed as individual difference variables belonging to certain people. These people, most likely men, will have power over others and may be more able to exert sexually harassing behaviors on those without power. These men may be more likely to abuse their powers if they possess the individual dispositional characteristics described earlier in the discussion of Pryor’s Likelihood to Sexually

Harass scale. The social and feminist models are also applicable here because both assert that men (in society and in relation to women) have more power. It is more likely, therefore, that men will be in possession of the different powers outlined in French and

Raven’s (1959) model of power.

These different types of power are also reflected in Ragins and Sundstrom’s

(1989) approach. Ragins and Sundstrom describe four different classifications of power 32 reflecting different levels that overlap with French and Raven’s sources of power.

Ragins and Sundstrom’s classifications include position power, interpersonal relationships, individual sources of power, and combinations of sources of power.

Position power overlaps nicely with legitimate, reward, and coercive powers such that these powers all stem from one’s position within a power hierarchy. Those at the top of the hierarchy will possess more of each of these powers. Legitimate, reward, and coercive powers can be difficult to tease apart as separate powers, however all three powers are usually a function of position power. The second category, interpersonal relationships, refers to powers that deal with “direction of influence” (p. 52). Direction of influence is illustrated well by Ragins and Sundstrom’s example of a supervisor and subordinate having a downward direction of influence such that the supervisor’s power flows down through the organization to the subordinate. Power stemming from interpersonal relationships can also be defined in terms of identification, such as French and Raven’s referent power, in which one individual strives to be like another.

Ragins and Sundstrom’s (1989) third classification of power is individual sources of power which refer to characteristics of an individual. Expert and informational power both fall under this category because these are sources of power that an individual possesses; someone has a large amount of knowledge or information desired or needed by others, thus this person has these sources of power.

The final classification of power is the combinations of sources of power.

Individuals may possess more than one of these powers at any given time. Ragins and

Sundstrom (1989) describe three hypotheses regarding the effects of combinations of powers. The first states that the more sources of power an individual possesses, the more 33 powerful he or she is; the powers are additive. It may also be, however, that the powers overlap, thus more sources do not necessarily equal more power. The second hypothesis is that the presence of one source of power can compensate for the lack of another.

Therefore, although the number of sources of power may remain constant, the loss of one source and the acquisition of another may result in no net change in influence. For instance, a manager may lose referent power over his co-workers when he gains a promotion, but due to his newly-acquired legitimate power, he can still influence his employees. The additive and compensatory hypotheses are very similar in that both hypotheses are concerned with the effects of having multiple sources of power. For example, compensatory sources of power may also be considered to be additive, in that the presence of one source of power can make up for the lack of another. The third hypothesis states that sources of power an individual has accumulated may gain or lose importance over time such that although the number sources stay the same, their usefulness may change as the organization evolves.

Other researchers have also assessed individual differences in power. Ragins and

Sundstrom (1989) reviewed some of the individual factors that may influence the amount of power a person can have. Some of these background factors include education and the presence of powerful role models. Other individual factors influencing individual power include one’s skills and abilities, personality traits such as a need for power and achievement orientation, and career aspirations. Finally, the presence and type of non- work roles can also impact individual power. For instance, many women may not be able to climb higher in the power hierarchy because of their non-work roles of mother and 34 homemaker whereas men may not have these responsibilities away from the job which allows them to focus on their careers.

Cleveland and Kerst (1993) have a slightly different view of what they call interpersonal/personal power. This form of power is characterized by examining the behaviors of individuals in power and how their subordinates react to the behaviors. The authors describe interpersonal power as being characterized by the perceived freedom that the individual with less power has to react to the individual with greater power. For instance, if there is a large gap between individuals with great power and those with little power, the individuals with little power will feel that they have little room to respond to the power-holder; they may feel that what the power-holder says, goes.

Finally, power is a key part of the feminist model because, as previously stated, one of the fundamental underlying assumptions is that there is a power differential between men and women, with men holding more power. Feminists attempt to raise the power status of women to put a stop to men’s abuse of their power. One of these abuses of power results in the sexual harassment of women.

Summary. Power can appear in its traditional form in sexual harassment where a powerful man is harassing a less powerful woman, or it can appear in more non- traditional forms of sexual harassment, such as same-sex sexual harassment (Stockdale,

Visio, & Batra, 1999) and the harassment of men (Berdhal, Magley, & Waldo, 1996).

Understanding both traditional and non-traditional forms of sexual harassment is important to understand sexual harassment as a whole. 35 Traditional and non-traditional sexual harassment Sexual harassment incidents vary from situation to situation. Because no two people are the same, it is not surprising that no two incidents of sexual harassment are going to be exactly the same in characteristics or outcomes. Although there are a number of possible forms of sexual harassment in the workplace, two major characteristics appear to be predominate: A man is the harasser of a woman, and the man likely has some form of power over the woman (Wilson & Thompson, 2001). These types of images are much more frequent than other non-traditional forms. There are, however, more non-traditional forms of sexual harassment. Some of the traditional and non-traditional forms of sexual harassment will be reviewed here to help with understanding the variety of ways sexual harassment can manifest itself.

Traditional sexual harassment. As previously stated, the most traditional form of sexual harassment involves a man in a powerful position harassing a woman in a position of little or no power. The 1995 U.S. Merit Systems Protection Board survey addressed the question of who appear to be the harassers. The survey showed that while a powerful man harassing a powerless woman was the most common conception of sexual harassment, it is much more common for victims to be harassed by co-workers.

Specifically, 79% and 77% of male and female (respectively) victims were not harassed by an individual with more organizational power. When taking into account past surveys conducted by the U.S. Merit Systems Protection Board, these statistics have remained relatively stable over time since the first survey in 1980.

Research also supports the finding that co-worker sexual harassment is more common than supervisor sexual harassment. Popovich, Licata, Nokovich, Martelli, & 36 Zoloty (1986) conducted two studies to examine the incidence and prevalence of sexual harassment behaviors. The results of these studies showed that while behaviors perpetrated by co-workers were more common, behaviors perpetrated by supervisors were perceived to be more serious.

Although supervisor sexual harassment is less common than co-worker sexual harassment, the perceived severity of this behavior is important to note because it shows how the element of power is tied to sexual harassment. One well-accepted notion of sexual harassment is that it is often present in situations of differential levels of power between men and women. A number of studies and theoretical writings have examined the issue of power and have found a link between differences in power and sexual harassment (e.g. Bargh, Raymond, Pryor, & Strack, 1995; and Wayne, 2000).

Bargh et al. (1995) were interested in investigating an automatic link between power and sex in the minds of men who scored highly on the Likelihood to Sexually

Harass (LSH) scale. Participants were primed with power-related, sex-related, or neutral words. The same power- and sex-related words were used as stimuli, and participants were to pronounce the stimulus word as soon as they could after it appeared on the screen. When power or sex-related terms were used as primes, men who scored highly on the LSH scale responded to the sex-related or power target words, respectively, more quickly. Using this subliminal priming task, the experimenters were able to show a link between power and sex for men with high scores on the LSH scale.

Perceptions of sexual harassment behaviors do not always match the reality of experienced sexual harassment. Even so, most researchers would likely agree that the above descriptions of sexual harassment incidents represent the more traditional form. 37 Although much research exists on the topic of traditional sexual harassment, the role of power is still not completely understood, especially in the more non-traditional forms of sexual harassment.

Contrapower sexual harassment. As previously mentioned, there is one unique form of sexual harassment, contrapower sexual harassment, which does not clearly fit into either the traditional legal guidelines or perceptual definitions of sexual harassment.

Examination of the EEOC guidelines does not make apparent that contrapower forms of harassment exist. Furthermore, many people may be unfamiliar with how power relationships can affect sexual harassment, thus decreasing the number of people who will perceive contrapower to be a valid form of sexual harassment. For example, in the formal sense, a subordinate harassing his/her supervisor qualifies as contrapower harassment. However, in the “informal” sense, a woman harassing a man can be considered contrapower because, socially and physically, men are considered to have more power. This last point may be debatable because some individuals may feel that women do hold “informal” power over men. For example, women have sexual power over men, such that a man may feel that if he does not comply with a woman’s wishes, she could make false harassment claims against him. When confronted with sexually harassing behaviors, it is also possible that some men may become conflicted about how to respond. For instance, a man may know that it is not appropriate to engage in the behavior being proposed to him, but still be excited at the prospect of the behavior and flattered to have been approached in the first place.

Contrapower sexual harassment is a non-traditional form of harassment referring to situations in which the victim has more power than the harasser. Benson (1984) coined 38 the term “contrapower” sexual harassment in discussing the sexual harassment of female professors by male students. Benson noted that, at the time, this type of sexual harassment was often ignored in research in academic settings.

Rospenda, Richman, and Nawyn (1998) state that when considering contrapower sexual harassment, traditional models of sexual harassment and power do not quite explain behavior. The authors mention that contrapower sexual harassment does not merely stem from organizational position power, but also from societal power. They cite

Benson (1984) as proposing this concept of the societal aspect of contrapower harassment, which makes it more difficult for organizations to understand this type of harassment when using established models. For instance, a female supervisor who is being harassed by a subordinate man may not be perceived as an incident of sexual harassment by the organization because the woman has power over the man. What this organization is neglecting to realize, however, is the idea that men are perceived to have more power in general from a societal view. Therefore, the male harasser in this incident is operating from a different basis of power than what the organization recognizes.

Rospenda and colleagues note that while models such as the societal model and sex-role spillover model can help to explain subordinate men harassing powerful women, the models cannot account for other forms of non-traditional harassment such as women harassing men and same-sex sexual harassment. This finding further demonstrates the lack of understanding of the non-traditional forms of harassment. Non-traditional sexual harassment can be explained, however, if one considers other sources of power (i.e., not only formal position power). Perhaps sources of power are not merely co-existing, but are even competing: A harasser’s sex (i.e., being male) becomes the more salient source of 39 power over a victim’s position power (i.e., the female victim is higher in the organizational hierarchy).

More recently, researchers have begun examining a greater variety of types of sexual harassment, particularly contrapower harassment. Matchen and DeSouza (2000) studied contrapower sexual harassment in academia by examining the behaviors of students directed at faculty members. The results of their survey demonstrated that female faculty members were the recipients of more sexually harassing behaviors from students than were male faculty members and, furthermore, the female faculty members were more bothered by the students’ behaviors. The authors discussed the shift of power in student-faculty relationships. While the faculty members have positional power, the authors state that the students also have power over the faculty stemming from anonymity. Specifically, students are able to write sexual comments on evaluations or place obscene phone calls without revealing their identities. Faculty evaluations can be especially problematic, because individuals with more power than the faculty members have access to the comments. Because these evaluations are considered in faculty career advancement, student harassment via the evaluations can be used as a form of retribution against faculty members and can have a large impact on faculty careers.

Other results of Matchen and DeSouza’s (2000) study revealed that 63% of the students surveyed reported committing at least one harassing behavior toward a faculty member and 53% of male and female faculty members reported experiencing harassing behaviors. These numbers also demonstrate the high frequency with which sexual harassment can occur in academic settings. 40 DeSouza and Fansler (2003) conducted a similar study to the Matchen and

DeSouza (2000) research. This survey revealed that about one-third of students reported sexually harassing faculty members, and that male students were more likely to sexually harass than were female students. According to the faculty survey, more than half of the male and female respondents reported being sexually harassed by students and, similar to the Matchen and DeSouza’s results, female faculty members were more bothered by the behavior than were male faculty members. DeSouza and Fansler’s study also demonstrated the relatively high prevalence of contrapower sexual harassment in academia.

Wayne (2000) investigated the relationships among a variety of variables of a harasser and the victim of harassment. Wayne manipulated several characteristics of the two individuals including; harasser gender, position power, harasser age (operationalized as either older or younger than the victim), and target age (operationalized as either older or younger than the harasser). The harasser was positioned as the victim's supervisor, co- worker, or subordinate, and measures on a variety of dependent variables were assessed.

Participants read a court case of a sexual harassment incident which was based, in part, on an actual case.

Across each of the dependent variables (verdict, unwelcomeness of the behavior, and organizational responsibility), subordinate-to-supervisor harassment was rated the most negatively. Subordinate harassers (i.e., subordinates who harassed someone with more formal power) were more likely to receive guilty verdicts, and their behavior was viewed as more unwelcome resulting in greater findings of organizational responsibility. 41 In Wayne’s (2000) research, harassing behaviors by older male harassers and younger female harassers were perceived by participants to be more welcome. Female participants tended to view the organization as being more responsible than did the male participants, although there were no participant gender differences on the verdict or unwelcomeness dependent variables.

Some of Wayne’ (2000) results were contrary to her hypotheses, including her hypothesis that subordinate harassers may be evaluated more positively than supervisor and coworker harassers. She did predict, however, that because of role-discrepant behavior, subordinate harassers may be evaluated negatively. Wayne (2000) offered a few possible explanations for her finding. First, she mentioned that organizational and societal theories of power only offer predictions of how individuals will perceive contrapower sexual harassment. Because contrapower is less often studied, there is not much information available to test these predictions. Although the results of the study contradicted the predictions, it is possible that the theory predictions are incorrect.

Second, Wayne mentioned that the surprising results of her study may simply be due to chance. This second explanation is one reason for the current study; to see if the same pattern of results occurs with a different sample. Third, Wayne proposed that the role- discrepant behavior of the subordinate harassers may have attracted attention resulting in more negative evaluations. Typically, a subordinate would not be thought of as asserting harassing behaviors and when put in this situation, any harassing behaviors become role- discrepant.

Contrapower sexual harassment also applies when examining the sexual harassment of men, although not in the same formal sense of the term as defined by 42 Benson (1984). Because men are often physically larger than women and because society often ascribes more powerful traits to men, situations in which men are the target of sexual harassment can be viewed as contrapower because the victim (the man) would seem to have more power over a female harasser. Even if the woman has more organizational power than her victim, the man still holds other bases of power over her

(e.g. physical size).

Summary. Contrapower sexual harassment has received considerably less attention in the research literature. Although this phenomenon has been studied (e.g.,

Wayne, 2000), it has not been completely understood and the many ways that contrapower can present itself have not been fully explored.

Although contrapower sexual harassment has been studied in both organizational and academic settings, based on previous research (including the studies reviewed here), contrapower harassment appears to be more prevalent in academic settings. Many characteristics distinguish academic from other organizational settings, some of which may influence the forms of sexual harassment that surface in each environment.

Organizational vs. Academic Settings Ilies, Hauserman, Schwochau, and Stibal (2003) conducted a meta-analysis examining the link between power and sexual harassment by categorizing studies based on the environment in which the study was conducted. The four resulting categories were; academic, private sector, government, and military. Their results showed that military settings experienced the highest incidence rate of sexual harassment, regardless of the survey instrument used. Academic settings had the lowest incidence rate when considering direct query surveys (as hypothesized by the researchers). However, when 43 examining behavioral experiences, academic settings had the second-highest incidence rate, higher than both private sector and government settings. The results of this meta- analysis reveal that, contrary to the authors’ predictions, women in academia are experiencing sexual harassment behaviors more often than women in other (non-military) occupations. Ilies and colleagues report that although the incidence rates were higher for women in academic settings, the actual report rates for the behaviors are lower in academia. The authors speculate that perhaps women in academia feel less threatened by sexual harassment, but agree that more research is needed in this area. Certainly, it is possible, however, that women in academia may feel more threatened by sexual harassment; thus, these women choose not to report harassing behaviors.

Contrapower sexual harassment, as well as more traditional forms of sexual harassment, can be more prevalent in academia, as reflected in Benson’s (1984) descriptive definition. Although academic and organizational settings are comparable, they are also different. The academic setting allows for a more detailed examination of how position power and gender can play into sexual harassment. In the typical organization, because there are so few women in supervisory positions, there is less opportunity for contrapower harassment. In a university, however, there are often female professors or women holding other positions of authority over a large number of students.

The supervisor in an organizational setting can be compared to the professor in an academic setting and the organizational subordinate is much like a student. It is easy for one to imagine a case where a male professor uses his power to extract sexual favors from his female students, but what about male students harassing female professors; or female students harassing male professors? A number of researchers have examined 44 sexual harassment in academia (e.g., Matchen & DeSouza, 2000; Gerrity, 2000;

Fitzgerald, Weitzman, Gold, & Ormerod, 1988) to investigate how particular aspects of the academic environment are affected by sexual harassment (e.g., examining how university employees, as opposed to students, are affected by harassment).

Fitzgerald, Weitzman, Gold, and Ormerod (1988) conducted a study to examine the relationships between male faculty members and their students in an attempt to investigate sexual harassment from the harasser’s viewpoint. The results of their survey indicated that over 25% of the respondents reported participating in sexual relationships with students in the same university, but that it does not appear to be related to characteristics such as the faculty member’s age or rank within the university. The authors believe that this number is likely an underestimate. Participants also indicated some of the characteristics that made faculty-student relationships permissible, including

(among others): whether the relationship was based on mutual consent, whether the faculty member would ever evaluate the student, and whether the student initiated the relationship. Fitzgerald and colleagues noted that within each of the reasons for justifying relationships with students, there was an element of power involved. For example, while it may not have been explicit, the faculty member has a position of power over the student which raises the question of whether there can really be mutual consent in beginning a relationship. This could be considered to be an extreme viewpoint, however, because one could question whether there could ever be a relationship without some sort of power differential between the parties involved. While demonstrating that sexual interactions do occur between faculty members and students, the results of Fitzgerald, et 45 al.’s study also indicate that these harassers (as Fitzgerald and her colleagues label these men) have many ways of justifying their actions.

Reilly, Carpenter, Dull, and Bartlett (1982) also examined sexual harassment in an academic setting. They mention that sexual harassment seems to differ between academic and organizational settings, primarily based on “the unique power relationship which exists between a student and her instructor” (pp. 100). Another key characteristic of the difference between academic and organizational settings that the authors mention is time. Students are required to interact with an instructor for a much smaller span of time than do individuals in other workplaces. Specifically, a student can drop a class or switch majors to leave the influence of a harasser (though students are also dependent upon faculty for future careers in terms of letters of recommendation, etc). In contrast, a woman in an organization will likely be required to remain in contact with her harasser for much longer. She may not be able to leave her job or may not feel comfortable making accusations against her superiors.

A third differentiating factor between academic and organizational sexual harassment described by Reilly, et al. (1982) is the age of the victim. Academic sexual harassment victims are generally young and inexperienced and rely on their professors for guidance. The organizational victim will likely be older and more experienced. The authors mention that sexual harassment will likely be perceived quite differently by these different-aged victims. Specifically, the working woman will be likely to perceive the behavior more negatively than would the student because she has more experience in the working world and understands that the way she is being treated is inappropriate. 46 Summary. Research has demonstrated that sexual interactions between faculty members and students do occur, and that they may occur at rates that exceed what has been reported (Fitzgerald, Weitzman, Gold, & Ormerod, 1988). Furthermore, though some of these interactions may be consensual, others certainly qualify as sexual harassment of the student by the faculty member. Research has also described differences between organizational sexual harassment and academic sexual harassment in terms of characteristics and frequency. The majority of the research has focused on male harassers and female victims. More research is needed to investigate the other combinations of the genders of the victims and harassers involved, especially female harassers and male victims. Related to this issue, research is also needed to further examine incidents of sexual harassment in which the harasser has, or appears to have, less power (biological, organizational, social, etc.) than the victim. Given that the incidence rate of sexual harassment in academic settings is somewhat higher than that in other organizations, and given that academic settings offer varying levels of power including many men and women in positions of high and low power, the academic setting is ideal for studying sexual harassment, especially contrapower sexual harassment.

The present study Building on previous research, especially Wayne (2000), the purpose of the current study was to examine contrapower sexual harassment in an attempt to further understand the role of power in both traditional and non-traditional forms of sexual harassment. Although contrapower sexual harassment has been explored, there have been requests in the research literature for a more detailed description of the relationship between power and sexual harassment (e.g., Wayne, 2000; Stockdale, et al., 1999). 47 Researchers have also suggested that some studies “have confounded position power with socio-cultural elements of power such as gender and age” (Wayne, 2000, pp. 303).

Perhaps instead of understanding the manipulation of position power in previous research studies, participants are merely reading to the gender or age of the harasser or of the victim and are making judgments based on these variables. It is easy to imagine how a participant might not have noticed that, in a given vignette, a woman is the supervisor of a man. The participant may base evaluations on the readily available idea of the man being the individual in charge and may ignore inconsistent information. In fact,

Popovich, Jolton, Mastrangelo, Everton, and Somers (1995) revealed that the script for a typical sexual harassment incident included a male supervisor harassing a female subordinate. Giving the individuals titles, such as “professor” and “student” should make the power manipulations more salient to the reader.

The current study used a methodology similar to that of Wayne (2000) and Reilly, et al. (1982) to investigate perceptions of various sexual harassment vignettes in which key elements of the interaction related to power were varied, including; harasser and victim gender, position power, as well as type of vignette. The study dealt with an instance of sexual harassment that both male and female participants would be likely to perceive as harassment because previous studies revealed participant gender differences in evaluations of less serious behaviors as sexual harassment. Therefore, the vignettes either portrayed such an instance of sexual harassment or no sexual harassment at all.

Instead of working within an organizational (e.g., court) setting as Wayne (2000) did, this study used an academic setting with professors and students as the harassers/victims. This choice was based on previous research which demonstrates the high prevalence of both 48 traditional and non-traditional forms of sexual harassment in academic settings (e.g.,

Matchen & DeSouza, 2000; Gerrity, 2000; Fitzgerald, Weitzman, Gold, & Ormerod,

1988). This research further examined the perceptual nature of sexual harassment in order to understand the dynamics of power in sexual harassment.

Hypotheses The current study had several hypotheses involving combinations of the variables of interest. The hypotheses for the study were as follows:

Hypothesis 1: It was hypothesized that there would be no differences between

any of the control vignettes, and that all of the control conditions would be

perceived more favorably (i.e., not harassment) than would the harassment

vignettes.

Hypothesis 2: Because of a greater familiarity with traditional forms of sexual

harassment, more vignettes with professors as harassers would be perceived to be

sexual harassment than would vignettes with students as harassers.

Hypothesis 3: Based on previous research showing different evaluations of

supervisor and subordinate harassers, and specifically that supervisory harassers

tend to be evaluated more negatively than other harassers (e.g., Popovich, et al.,

1986), it was predicted that professor harassers would be evaluated more

negatively than would student harassers.

Hypothesis 4: Because men, and not women, are generally viewed as the

instigators of sexual harassment, it was hypothesized that male harassers would be

evaluated more negatively than would female harassers (regardless of position

and age). 49 Hypothesis 5: This hypothesis refers only to the contrapower conditions, either

formal (i.e., a student harassing a professor) or “informal” (i.e., based on some

other characteristic such as social privilege). With respect to the contrapower

conditions, it was predicted that the conditions where the harasser was female

(either student or professor), as well as the condition where the male harasser

was a student, would generally be evaluated more negatively than the more

traditional form of harassment displayed in the condition where the harasser was a

supervisory male over a female victim. This hypothesis was based on Wayne’s

(2000) findings where results showed that female harassers were perceived to be

more guilty, as well as her finding that subordinate harassers were perceived to be

more guilty, more unwelcome, and resulted in findings of more organizational

responsibility. It has been previously hypothesized that male harassers and

professor harassers will be evaluated more negatively (see Hypotheses 1 and 2),

however, when combining these two factors, it was predicted that the less often

recognized details of the vignettes (i.e., the non-traditional and contrapower

aspects) would cause the pattern of means to switch direction and result in more

negative evaluations of these groups.

Hypothesis 6: Because of their greater familiarity with sexual harassment, participants reporting previous sexual harassment training would be more likely to

perceive the harassment vignettes as portraying sexual harassment.

Hypothesis 7: Because previous meta-analyses that have shown small, if any,

gender differences regarding less ambiguous forms of sexual harassment (e.g.,

quid pro quo), it was hypothesized that male and female participants would show 50 no significant mean differences on their determinations of whether sexual

harassment had occurred in the vignettes.

Method

Participants Participants were 282 male (42.9%) and female (55.7%) undergraduates enrolled in introductory psychology classes with a mean age of 19.28 years. Participants were largely Caucasian (86.9%) and were representative of the various colleges at the university. Participants volunteered for the research via an online experiment sign-up webpage and were be given research points for their participation which could be applied toward grades in their psychology classes.

Design The study was a 2 (participant gender) x 2 (harasser gender) x 2 (victim gender) x

2 (harasser position power: supervisor to or subordinate of the victim) x 2 (“harassment” or no “harassment”) between-subjects factorial design, such that each participant read only 1 vignette. Participants read a short vignette in which these elements had been varied to create different settings in which various forms of sexual harassment, including contrapower sexual harassment, may arise (See Appendix B for vignettes). Because participants often have differing opinions regarding whether hostile environment incidents constitute sexual harassment (e.g., Blumenthal, 1998; Rotundo, Nguyen &

Sackett, 2001), the vignettes presented to participants contained incidents of sexual harassment that were more likely to be perceived as such or no sexual harassment, as a control. 51 Materials Participants were given a consent form explaining the study procedure as well as potential benefits and risks associated with participating in the research. Vignettes were provided for participants in which, as previously stated, key variables were manipulated to create unique circumstances. Participants’ perceptions regarding the vignette were assessed, as were more general perceptions of sexual harassment and demographic information. Some of the questions pertaining to the vignette included; the participants’ perceptions of whether the incident constituted sexual harassment, which party was to blame for the behavior, whether the organization was responsible for the harassment and, if so, to what extent, how unwelcome and/or bothersome the behavior was to the victim, and estimates of the ages of the individuals involved. Each question was accompanied by a scale ranging from 0 to 5 with the following anchors: 0- Not applicable; 1- Clearly not;

2- Probably not; 3- May or may not; 4- Probably; and 5- Clearly. The anchors were expanded for each question specifically. See Appendix C for the questionnaire items.

Because of Wayne’s (2000) statement that position power may be confounded with variables such as age, instead of manipulating this variable, participants were queried as to their estimations of the ages of the vignette characters to see where any perceived age differences may appear. Finally, participants were asked to supply the definition of sexual harassment that they personally used when making their evaluations of the vignettes as a response to an open-ended item.

Figure 1 presents the various conditions in the study. The conditions of most interest in the study were the ones reflecting contrapower sexual harassment. Those conditions included: a female student harasser of a male professor and a male student 52 harasser of a female professor. Although not formally contrapower, a female professor harasser of a male student is also of interest because it may be “informally” considered as a form of contrapower harassment. All of these conditions were compared to the most widely accepted image of sexual harassment consisting of a male professor harasser of a female student.

Figure 1. Vignette conditions included in the current study.

Harassment

Male Victim Female Victim

Subordinate Subordinate Supervisor Harasser Harasser Supervisor Harasser Harasser

non-traditional, non-traditional contrapower traditional contrapower

No Harassment

Male Victim Female Victim

Subordinate Subordinate Supervisor Harasser Harasser Supervisor Harasser Harasser

non-traditional, non-traditional contrapower traditional contrapower

Participants were also surveyed regarding their participation in sexual harassment

training, and more general demographic questions including gender, age, ethnicity, and

college major. 53 Procedure Upon arriving, participants were given a consent form to read and any questions regarding the experiment were answered. Participants were assigned to a condition based on when they signed up for the experiment, such that all participants in a given session were in the same condition. Near the end of data collection, conditions were assigned to achieve approximately equal sample sizes across conditions. Signed consent forms were collected by the experimenter and vignette/questionnaire packet was distributed to the participants. The experimenter reviewed the instructions for the packet with the participants and any questions were answered. Each participant read a vignette representing one of the 8 conditions of interest (4 sexual harassment and 4 no harassment conditions) and responded to the corresponding questionnaire items (see Appendix C).

Participants were given time to complete the packet which were then collected by the experimenter when finished. Any questions were again answered and participants were given credit, thanked, and dismissed.

Results The data collected in this study were coded in several ways in order to test the hypotheses. First, the data were recoded into a variable grouping the vignettes as either

“harassment” or “no harassment”: Vignettes 1 through 4 were classified as “no harassment” and vignettes 5 through 8 were classified as “harassment”. Second, the vignettes were recoded into a variable designating whether the vignette contained a professor harasser/initiator or a student harasser/initiator. Note that the term “initiator” is used to refer to the character in the “no harassment” conditions who initiates an interaction, however the interaction does not contain harassing behaviors. Vignettes 1, 3, 54 5, and 7 portrayed a professor harasser/initiator, while 2, 4, 6, and 8 portrayed a student harasser/initiator.

As will be explained, the results of Hypothesis 1 examining the extent of harassment perceived in the “harassment” versus “no harassment” indicated that the

“harassment” conditions were perceived to portray sexual harassment to a greater extent than did the “no harassment” conditions. Therefore, the main analyses were conducted using only the “harassment” vignettes. Because the main hypotheses only dealt with the

“harassment” conditions, the professor/student variable was coded to only include the

“harassment” vignettes. Finally, a variable was created in order to group male and female harassers/initiators: Vignettes 3, 4, 7, and 8 portrayed male harassers/initiators, while 1,

2, 5, and 6 portrayed female harassers/initiators. Again, in analyzing the main hypotheses, this variable was recoded to contain only the “harassment” vignettes.

Based on examination of the means for questions 1 through 6, the participants in each of the different conditions answered as would be expected for the condition, indicating that these questions function as manipulation checks for the different conditions. For example, in vignettes containing a male harasser, the means were higher for the extent to which the female character was harassed (e.g., 4.69 for the male professor harasser versus 1.00 for the female professor harasser), the extent to which the behavior was intentional on the part of the male character (e.g., 4.80 for the male professor harasser versus 1.23 for the female professor harasser), and the extent to which the male character was to blame for the behavior (e.g., 4.71 for the male professor harasser versus 1.49 for the female professor harasser). Therefore, it appears that the 55 manipulations of harasser position power and harasser gender were successful. See

Appendix D for a complete list of manipulation means.

A series of multivariate analyses of variance (MANOVA) were conducted to examine the hypotheses. Open-ended questions were also content analyzed to look for any patterns in the responses.

Hypothesis 1

It was hypothesized that the “no harassment” conditions would be perceived as containing less harassment than would the “harassment” conditions, and that there would be no significant differences regarding this hypothesis between the four “no harassment” conditions. Table 1 displays the means and test statistics for this analysis. Two separate

MANOVAs were conducted to investigate various differences between the “harassment” and the “no harassment” vignettes on questions 1 and 2, concerning the extent to which harassment was perceived to have occurred. The multivariate test of this hypothesis was significant, F(2,279)= 183.05, p<.001. As hypothesized, for both the male and female victims (F(1,280)= 69.59, p<.001 and F(1,280)= 91.68, p<.001, respectively), the harassment vignettes were considered to show a greater degree of harassment than were the no harassment vignettes.

56 Table 1. Multivariate Analysis of Variance and Means for Type of Vignette. Source F Multivariate tests Vignette Wilks’ Lambda 183.05*** Type (2,279)=.43 Between subjects H NH X (SD) X (SD) Vignette Male Character 2.52 1.20 69.59*** Type Harassed (1.75) (0.71) Female Character 2.86 1.20 91.68*** Harassed (1.92) (0.73) Note. Univariate df=1, 280. H= Harassment, NH= No Harassment. Harassment N= 140; No Harassment N=142. *p<.05, **p<.01, ***p<.001

A second MANOVA was conducted to look for any differences between the control vignettes. Table 2 displays the means and test statistics for this analysis. The multivariate test was significant, F(6,274)= 4.75, p<.001. In partial support of Hypothesis

1, there were no significant differences between any of the control conditions for whether the male victim was harassed, F(3,138)= 1.07, p>.05. The male professor initiator

vignette was perceived to contain significantly more harassment than any of the other

control vignettes, F(3,138)= 6.86, p<.001, indicating that perhaps participants were

slightly more likely to perceive sexual harassment because of this traditional power

relationship. It is important to note, however, that the mean value for this condition was

less than 2, indicating that the participants did not perceive the vignette as portraying

sexual harassment. The remaining 3 control vignettes showed no differences on this

variable.

57

Table 2. Multivariate Analysis of Variance and Means for Type of Vignette, Control Vignettes Only. Source F Multivariate tests Control 4.75*** Vignettes Wilks’ Lambda (6,274)=.82 Between subjects MSI FSI MPI FPI X (SD) X (SD) X (SD) X (SD) Control Male Character 1.11 1.11 1.20 1.37 1.07 Vignettes Harassed (0.57) (0.67) (0.90) (0.65) Female Character 1.11 1.06 1.66 1.00 6.86*** Harassed (0.67) (0.53) (1.00) (0.73) Note. Univariate df=3,138. MSI= Male Student Initiator, FSI= Female Student Initiator, MPI= Male Professor Initiator, FPI= Female Professor Initiator. Male and Female Student Initiator N=36, Male and Female Professor Initiator N=35. *p<.05, **p<.01, ***p<.001

Hypothesis 2

This hypothesis contained the prediction that, because of a greater familiarity with

the traditional forms of sexual harassment, the vignettes with professors as harassers

would be perceived to be sexual harassment to a greater extent than would vignettes with

students as harassers. Table 3 displays the test statistics and means for this analysis. A

MANOVA was conducted comparing professor and student harassers on questions 1 and

2. The multivariate test of Hypothesis 2 was significant, F(2,137)= 3.83, p<.05. In

support of the hypothesis, professor harasser vignettes were perceived to portray

harassment to a greater extent for the male victim than were student harasser vignettes,

F(1,138)= 3.99, p<.05. 58 Contrary to the hypothesis, however, there were no significant differences between professor and student harasser vignettes for the female victim (F(1,138)= .01, p>.05). This indicates that the female professor harasser was evaluated more negatively than was the female student harasser. The male professor and student harassers were perceived to be equally harassing.

Table 3. Multivariate Analysis of Variance and Means for Type of Harasser. Source F Multivariate tests Harasser Type Wilks’ Lambda (2,137)=.95 3.83* Between subjects Student Professor X (SD) X (SD) Harasser Type Male Character Harassed 2.23 (1.60) 2.81 (1.86) 3.99* Female Character Harassed 2.87 (1.92) 2.84 (1.94) 0.01 Note. Univariate df=1, 138. Student N=70; Professor N=70. *p<.05, **p<.01, ***p<.001

Hypothesis 3

Because supervisory harassers tend to be evaluated more negatively than

subordinate harassers, it was hypothesized that professor harassers would be evaluated

more negatively than would student harassers. A MANOVA was conducted to examine differences between the two groups on questions 3 through 27. These questions included; the extent to which the behavior was intentional, who was to blame (including the university), the extent to which the behavior was perceived to be unwelcome or bothersome, the extent to which there was an abuse of power, the extent to which there 59 was a fear of personal repercussions, the extent to which the behavior created an intimidating, hostile, or offensive classroom environment, and finally, the perceived morality and attractiveness of the characters. Table 4 displays the test statistics for this analysis. See Appendix E for means. The multivariate test of Hypothesis 3 was significant, F(25,113)= 3.58, p<.001.

Of the many questions included, only three variables showed significant differences between professor and student harassers: The extent to which the university was to blame (F(1,137)= 22.80, p<.001); the extent to which the female character abused her power (F(1,137)= 7.79, p<.005); and how ethical/moral the female character was perceived to be (F(1,137)= 7.37, p<.006). Specifically, for vignettes portraying a professor harasser, participants felt that the university was to blame to a greater extent than when the vignettes portrayed a student harasser. Also, the female professor harasser was perceived to have abused her power to a greater extent than did the female student harasser. Finally, participants evaluating vignettes containing student harassers perceived the female character to be more ethical/moral than did participants evaluating professor harassers.

Because there appeared to be relatively few professor/student differences,

Hypothesis 3 was largely unsupported. However, three variables showed results in the predicted direction: The extent to which the university was to blame; the extent to which the female character abused her power; and the extent to which the female character was ethical. Specifically, these three results indicate that professor harassers may be evaluated more negatively than would student harassers.

60 Table 4. Multivariate Analysis of Variance for Professor vs. Student Harassers. Source F Multivariate tests Harasser Type Wilks’ Lambda (25, 133)=.56 3.58*** Between subjects Harasser Type Female Acted With Intent 1.30 Male Acted With Intent 0.05 Female to Blame 0.11 Male to Blame 0.09 University to Blame 22.80*** Male Found Behavior Unwelcome 0.68 Female Found Behavior Unwelcome 1.05 Female Abused Power 7.79** Male Abused Power 0.17 Male Feared Personal Repercussions 0.65 Female Feared Personal Repercussions 0.12 Pretend to be Male Character 0.00 Pretend to be Female Character 0.02 Female Interfered With Environment 0.17 Male Interfered With Environment 0.12 Female Created Intimidating Environment 0.79 Male Created Intimidating Environment 0.16 Female Created Hostile Environment 2.20 Male Created Hostile Environment 1.16 Female Created Offensive Environment 0.48 Male Created Offensive Environment 0.10 Female’s Physical Attractiveness 3.69 Male’s Physical Attractiveness 1.94 Female’s Ethics/Morals 7.37** Male’s Ethics/Morals 1.72 Note. Univariate df=1, 137. Student N=70, Professor N=69. *p<.05, **p<.01, ***p<.001

An additional MANOVA was conducted to further break down the differences

between the professor and student harassers. This analysis compared each of the

“harassment” vignettes on questions 1 through 27 to check for individual differences

between the “harassment” vignettes individually. Table 5 displays the test statistics for 61 this analysis. See Appendix E for means. Specifically, this analysis was conducted to determine whether the differences were more pronounced in particular vignettes (e.g., the non-contrapower vignettes).The multivariate test of this analysis was significant,

F(81,326.88)= 12.97, p<.001.

Nearly all of the variables revealed significant differences among the vignettes at the p<.001 level, including: The extent to which each character was harassed; the extent to which the behavior was intentional; the extent to which each character, and the university, was to blame; the extent to which the behavior was perceived to be unwelcome; the extent to which there was an abuse of power; the extent to which there was a fear of personal repercussions; how uncomfortable the participants would be with the behavior, were they one of the characters; the extent to which the behavior created an intimidating, offensive, or hostile environment; the male character’s physical attractiveness; and the female character’s morals. The two variables showing non- significant differences were the physical attractiveness of the female character, and the extent to which the male character was ethical or moral.

Because this analysis was intended to investigate whether the non-contrapower conditions were different from the contrapower conditions, Tukey post-hoc tests were conducted. Four variables had significant differences, indicating that the non- contrapower conditions were evaluated more negatively than were the contrapower conditions. All other variables revealed no differences, indicating that contrapower versus non-contrapower differences were not present.

Specifically, the female professor harasser was perceived to have acted with intent to a greater extent than was the female student harasser. Participants evaluating vignettes 62 containing both the male and female professor harassers perceived the university to be to blame to a greater extent than did participants evaluating male and female student harasser vignettes.

The female professor was perceived to have abused her power to a greater extent than did the female student harasser, and the male professor harasser was perceived to have abused his power to a greater extent than did the male student harasser.

63 Table 5. Multivariate Analysis of Variance for Harassment Vignettes Only. Source F Multivariate tests Vignette Wilks’ Lambda (81,326.88)=.01 12.97*** Between subjects Vignette Male Harassed 341.83*** Female Harassed 61.08*** Female Acted With Intent 202.14*** Male Acted With Intent 344.87*** Female to Blame 238.48*** Male to Blame 244.15*** University to Blame 7.82*** Male Found Behavior Unwelcome 37.06*** Female Found Behavior Unwelcome 125.79*** Female Abused Power 75.35*** Male Abused Power 77.88*** Male Feared Personal Repercussions 38.50*** Female Feared Personal Repercussions 24.18*** Pretend to be Male Character 61.27*** Pretend to be Female Character 178.48*** Female Interfered With Environment 42.88*** Male Interfered With Environment 26.14*** Female Created Intimidating Environment 122.45*** Male Created Intimidating Environment 115.60*** Female Created Hostile Environment 38.54*** Male Created Hostile Environment 83.92*** Female Created Offensive Environment 85.56*** Male Created Offensive Environment 95.01*** Female’s Physical Attractiveness 1.49 Male’s Physical Attractiveness 7.89*** Female’s Ethics/Morals 3.49* Male’s Ethics/Morals 1.86 Note. Univariate df=3,135. *p<.05, **p<.01, ***p<.001

Hypothesis 4

This hypothesis contained the prediction that male harassers would be evaluated

more negatively than would female harassers. The same questionnaire items were 64 included in the analysis as were included in the previous analysis of Hypothesis 3, with the addition of the extent to which each individual was harassed (questions 1 and 2).

Table 6 displays test statistics for this analysis. See Appendix F for means. The multivariate test of these variables was significant, F(27,111)= 87.63, p<.001.

The following findings are as would be expected, given that all vignette interactions were heterosexual. Specifically, the male character was perceived to be harassed to a greater extent when the vignette portrayed a female harasser. Also, the female character was perceived to be harassed to a greater extent when the vignette portrayed a male harasser.

Similarly, the female character was perceived to have acted with intent to a greater extent for the female harasser vignettes, and the same is true for the male character in the male harasser vignettes as opposed to the female harasser vignettes. The female character was perceived to be to blame to a greater extent when the harasser was a female, and the male character was perceived to be to blame to a greater extent when the harasser was a male.

The unwelcomeness of the behavior was perceived to be greater for an individual when the harasser was of the opposite sex. Specifically, the male character was perceived to have found the behavior unwelcome to a greater extent when the harasser was a female, and the female character was perceived to have found the behavior more unwelcome when the harasser was a male.

The individual who was perceived to have abused his/her power to a greater extent was the individual matching the harasser’s gender. Specifically, the female character was perceived to have abused her power to a greater extent when the harasser 65 was a female, and the male character was perceived to have abused his power to a greater extent when the harasser was a male.

The male character was perceived to fear more personal repercussions when the vignette contained a female harasser rather than a male harasser. Also, the female character was perceived to fear more personal repercussions when the vignette contained a male harasser.

When asked to pretend to be the male character, participants reported that they would feel uncomfortable with the behavior of the female harasser to a greater extent.

Also, when asked to pretend to be the female character, participants reported that they would feel uncomfortable with the behavior of a male harasser to a greater extent.

In vignettes containing a female harasser, participants reported that they perceived the behavior to interfere with the male character’s work performance, as well as to create an intimidating, offensive, and hostile classroom environment for him. A similar pattern holds for vignettes containing a male harasser in that the behavior was perceived to interfere with the female character’s work performance, as well as to create an intimidating, offensive, and hostile classroom environment for her. Finally, participants perceived the male character to be more physically attractive for vignettes containing a female harasser.

There were no harasser gender differences for the extent to which the university was to blame, the female character’s physical attractiveness, or either question regarding the ethics of the vignette characters.

Rather than following a pattern suggesting that male harassers are evaluated more negatively than were female harassers across all variables (which would support the 66 relationship hypothesized in Hypothesis 4), it appears that the individual portrayed as the harasser was generally perceived more negatively, regardless of gender. Additional analyses revealed that, when considering the items concerning how harassed the victim was perceived to be, there were differences among the types of harassers

(F(3,136)=19.334, p<.001). The male professor harasser ( X =4.69) was perceived to be more harassing than the female professor ( X =4.17) and female student ( X =3.46)

harassers, but was equivalent to the male student harasser ( X =4.63). Furthermore, the

male student harasser was perceived to be more harassing than the female student

harasser, but equivalent to the female professor harasser. Finally, the female professor

harasser was perceived to be more harassing than the female student harasser.

67 Table 6. Multivariate Analysis of Variance for Male vs. Female Harassers. Source F Multivariate tests Harasser Type Wilks’ Lambda (27,111)= .05 87.63*** Between subjects Harasser Type Male Harassed 164.60*** Female Harassed 1035.38*** Female Acted With Intent 554.48*** Male Acted With Intent 1041.26*** Female to Blame 713.46*** Male to Blame 675.73*** University to Blame 0.68 Male Found Behavior Unwelcome 102.34*** Female Found Behavior Unwelcome 370.37*** Female Abused Power 158.16*** Male Abused Power 186.01*** Male Feared Personal Repercussions 107.80*** Female Feared Personal Repercussions 70.01*** Pretend to be Male Character 183.51*** Pretend to be Female Character 542.12*** Female Interfered With Environment 125.57*** Male Interfered With Environment 73.55*** Female Created Intimidating Environment 353.56*** Male Created Intimidating Environment 329.49*** Female Created Hostile Environment 106.45*** Male Created Hostile Environment 243.65*** Female Created Offensive Environment 253.15*** Male Created Offensive Environment 268.61*** Female’s Physical Attractiveness 0.82 Male’s Physical Attractiveness 20.60*** Female’s Ethics/Morals 0.13 Male’s Ethics/Morals 3.51 Note. Univariate df=1, 137. Male N=69, Female N=70. *p<.05, **p<.01, ***p<.001

Hypothesis 5

Based on prior research (Wayne, 2000), it was hypothesized that the two

conditions in which the harasser was a female (student and professor), as well as the 68 condition in which there was a student male harasser, would be evaluated more negatively than would the “traditional” condition portraying a male professor harasser. A

MANOVA was again conducted, and the interaction of harasser gender and harasser position power was assessed using questions 1 through 27. Table 7 displays test statistics for this analysis. See Appendix G for means. The multivariate test of this hypothesis was significant, F(27,109)= 3.16, p<.001.

Seven variables revealed significant interactions of harasser gender and harasser position power including: The extent to which the female character acted with intent

(F(1,135)= 5.81, p<.05); the extent to which the male character found the behavior to be unwelcome/bothersome (F(1,135)= 4.86, p<.05); the extent to which both the female and the male characters abused their power (F(1,135)= 14.05, p<.001 and F(1,135)= 20.76, p<.001, respectively); the extent to which the male character feared personal repercussions (F(1,135)= 4.16, p<.05); the extent to which the male character created an intimidating classroom environment (F(1,135)= 5.72, p<.05); and finally, the extent to which the male character created an offensive classroom environment (F(1,135)= 6.41, p<.05).

Specifically, there were no significant differences for the extent to which the female character acted with intent for vignettes containing male professor and student harassers. For vignettes containing female harassers, professor harasser vignettes received higher mean intent ratings than did student harasser vignettes.

When examining the extent to which the male character perceived the behavior to be unwelcome, the behavior of a male professor harasser was perceived to be more unwelcome than that of a male student harasser. For female harassers, however, the 69 pattern was reversed, and the behavior of a female student harasser was perceived to be more unwelcome for the male character.

The patterns for the extent to which the male and female characters abused their power differ. When considering the extent to which the female character abused her power, for vignettes containing male harassers, student and professor vignettes showed equivalent means, which were very low. Female professor harassers, however, were perceived to have abused their power to a greater extent than did female student harassers. When examining the extent to which the male character abused his power, male professor harassers were perceived to have abused their power to a greater extent than did male student harassers. For female harasser vignettes, however, the pattern was reversed, such that female student harassers were perceived to have abused their power to a greater extent than did female professor harassers, however the mean values were very low.

The extent to which the male character feared personal repercussions for vignettes containing male harassers was perceived such that he was more likely to fear repercussions for vignettes containing professor harassers than student harassers, however the mean values were low. The pattern was reversed for female harasser vignettes, such that the male was perceived to have feared personal repercussions to a greater extent stemming from the actions of the female student harasser.

The same pattern was present for the extent to which the male character created an intimidating and an offensive classroom environment. For vignettes containing male harassers, the male character was perceived to have created a more intimidating and offensive environment when the harasser was a professor rather than when the male 70 character was a student. This pattern was reversed, however, for vignettes containing female harassers. The male character in these vignettes was perceived to have created a more intimidating and offensive environment when the harasser was a student, however these means are very low.

Overall, for the variables with a significant interaction of harasser gender and harasser position power, Hypothesis 5 was partially supported. When examining vignettes with female harassers, five variables revealed that student harassers were evaluated more negatively: The extent to which the male character found the behavior to be unwelcome; the extent to which the male character abused his power; the extent to which the male character feared personal repercussions; the extent to which the male character created an intimidating and offensive environment. These results support previous research, showing that subordinate harassers may be evaluated more negatively.

For two variables, however, the professor harasser was perceived considerably more negatively than was the student harasser: The extent to which the female character acted with intent and abused her power. When considering the male harasser vignettes, the professor harasser was evaluated more negatively, or there were no differences between the professor and student harasser vignettes. These results do not support Hypothesis 5.

There were no other significant interactions for harasser gender and harasser position power.

71 Table 7. Multivariate Analysis of Variance for Harasser Gender x Position Power Interaction. Source F Multivariate tests Harasser Gender Wilks’ Lambda (27,109)=.56 3.16*** x Harasser Position Power Between subjects Gender x Power Male Harassed 0.39 Female Harassed 0.23 Female Acted With Intent 5.81* Male Acted With Intent 0.81 Female to Blame 1.60 Male to Blame 0.84 University to Blame 0.01 Male Found Behavior Unwelcome 4.86* Female Found Behavior Unwelcome 0.00 Female Abused Power 14.05*** Male Abused Power 20.76*** Male Feared Personal Repercussions 4.16* Female Feared Personal Repercussions 2.11 Pretend to be Male Character 1.26 Pretend to be Female Character 0.02 Female Interfered With Environment 2.14 Male Interfered With Environment 3.73 Female Created Intimidating 2.86 Environment Male Created Intimidating Environment 5.72* Female Created Hostile Environment 2.38 Male Created Hostile Environment 0.51 Female Created Offensive Environment 1.42 Male Created Offensive Environment 6.41* Female’s Physical Attractiveness 0.00 Male’s Physical Attractiveness 0.63 Female’s Ethics/Morals 2.92 Male’s Ethics/Morals 0.27 Note. Univariate df=1,135. Student N=35, Professor N=34. *p<.05, **p<.01, ***p<.001

72 Hypothesis 6

It was hypothesized that, due to an increased familiarity with sexual harassment, those participants reporting previous sexual harassment training will report the vignettes as containing more sexual harassment. A MANOVA was conducted to examine differences between these two groups on questions 1 and 2, concerning the extent to which harassment occurred. Table 8 displays the test statistics and means for this analysis. The multivariate test of this hypothesis was non-significant, F(4,556)= 2.30, p>.05. Contrary to the hypothesis, there were no differences in perceptions of whether harassment had occurred between those participants who have and those who have not had sexual harassment training on the questions regarding whether the male and female characters were harassed.

Table 8. Multivariate Analysis of Variance and Means for Participant Training Experience. Source F Multivariate tests Training Wilks’ Lambda (4,556)=.97 2.30 Between subjects Training No Training X (SD) X (SD) Training Male Character Harassed 1.60 (1.37) 1.94 (1.51) 1.51 Female Character Harassed 2.21 (1.69) 1.89 (1.62) 3.26 Note. Univariate df= 2, 279. Training N= 75; No Training N=193. *p<.05, **p<.01, ***p<.001

73 Hypothesis 7

Finally, it was hypothesized that, although female participants may be slightly more likely to evaluate the vignettes negatively, there would be no significant gender differences on questions 1 and 2. The data were split by vignette, and a MANOVA was conducted to investigate differences on these two questions. Tables 9 displays test statistics for this analysis. See Appendix H for means. Regardless of vignette type, none of the multivariate tests were significant. Consistent with the hypothesis, there were no significant participant gender differences on whether harassment occurred.

74 Table 9. Multivariate Analysis of Variance for Participant Gender. Source df F Multivariate tests Participant Male Student Initiator Wilks’ Lambda (2,31)=.21 1.21 Gender Female Student Initiator Wilks’ Lambda (2,33)=.99 0.02 Male Professor Initiator Wilks’ Lambda (2,31)=.85 2.65 Female Professor Initiator Wilks’ Lambda (2,32)=.90 1.75 Male Student Harasser Wilks’ Lambda (2,32)=.96 0.73 Female Student Harasser Wilks’ Lambda (2,32)=.90 1.74 Male Professor Harasser Wilks’ Lambda (2,31)=.85 2.75 Female Professor Harasser Wilks’ Lambda (2,32)=.92 1.45 Between subjects Variable Harasser Male Student Initiator Male Harassed 1 2.09 Gender Female Harassed 1 1.53 Error Male Harassed 32 Female Harassed 32 Female Student Initiator Male Harassed 1 0.03 Female Harassed 1 0.01 Error Male Harassed 34 Female Harassed 34 Male Professor Initiator Male Harassed 1 0.41 Female Harassed 1 4.57 Error Male Harassed 32 Female Harassed 32 Female Professor Initiator Male Harassed 1 0.47 Female Harassed 1 2.72 Error Male Harassed 33 Female Harassed 33 Male Student Harasser Male Harassed 1 0.89 Female Harassed 1 0.66 Error Male Harassed 33 Female Harassed 33 Female Student Harasser Male Harassed 1 3.06 Female Harassed 1 0.13 Error Male Harassed 33 Female Harassed 33 Male Professor Harasser Male Harassed 1 2.28 Female Harassed 1 3.08 Error Male Harassed 32 Female Harassed 32 Female Professor Harasser Male Harassed 1 2.99 Female Harassed 1 .00 75 Table 9: continued Error Male Harassed 33 Female Harassed 33 *p<.05, **p<.01, ***p<.001

Content of Open-Ended Questions

Content analysis is a procedure used to draw conclusions based on qualitative data

(Whitley, 2002). The process involves four general steps: 1) determine the source(s) of

the data; 2) take sample data from the source(s); 3) develop a coding scheme for the data;

and 4) determine how to measure the data. These steps applied to the current study are as follows: 1) the source of the data was the open-ended items included on the questionnaire; 2) the responses to these items were compiled and grouped by vignette; 3) the responses were coded for the presence of an abuse of power, the presence of inappropriate behaviors, the creation of an uncomfortable/threatening environment, and the perceived typicality of the student/professor interaction. Finally, 4) each response was coded as to whether or not it contained a reference to one of the coded categories.

Two open-ended items were included on the questionnaire, questions 30 and 31.1

Question 30 contains a request for participants to describe, in their own words, how they

decided whether the vignette contained sexual harassment. The responses were grouped

by vignette and were examined in their groups.

1 Question 31 contained a request for the participants to describe any sexual harassment training they had participated in (if any). Because of the non-significant differences between those who have and have not had training (see Hypothesis 6), this question was not analyzed. A review of the description of training provided by those who had received such training revealed that it was very limited and basic, often occurring in classes during high school 76 Table 10 contains categorizations of the responses. Four of the vignette conditions contained references to power/abuse of power. Both female initiators (student and professor) had one response suggesting an abuse of power. In the “harassment” conditions, both professor harassers (male and female) received several responses (7 and

5, respectively) suggesting an abuse of power. While only a few of the “no harassment” responses indicated the presence of inappropriate behaviors, between 54% and 74% of participants reading the “harassment” conditions made references to inappropriate behaviors occurring between the vignette characters. Similarly, while only a few responses in the “no harassment” conditions suggested that the behaviors of one of the parties created an uncomfortable or threatening environment, between 9% and 17% of the participants reading the “harassment” conditions specifically mentioned such an environment. Finally, 6% to 11% of participants who read the “no harassment” conditions described the vignettes as portraying nothing more than a typical interaction between a student and a professor. None of the “harassment” conditions received this endorsement.

These results indicate that, when comparing the “no harassment” to the

“harassment” conditions, the “harassment” conditions were perceived by participants as more likely to portray an abuse of power and uncomfortable/threatening environment due to inappropriate behaviors. The “no harassment” conditions, however, were more likely to be perceived as nothing more than a typical interaction between a student and professor, with very few indications that inappropriate behaviors or an abuse of power had taken place.

77 Table 10. Categorizations of Open-Ended Responses. Uncomfortable/ Inappropriate Threatening Vignette Power Behaviors Environment Typicality Female Professor Initiator, N=35 1 (3%) 0 (0%) 0 (0%) 4 (11%) Female Student Initiator, N=36 1 (3%) 2 (6%) 2 (6%) 2 (6%) Male Professor Initiator, N=35 0 (0%) 1 (3%) 1 (3%) 2 (6%) Male Student Initiator, N=36 0 (0%) 1 (3%) 0 (0%) 4 (11%) Female Professor Harasser, N=35 5 (14%) 20 (57%) 6 (17%) 0 (0%) Female Student Harasser, N=35 0 (0%) 19 (54%) 3 (9%) 0 (0%) Male Professor Harasser, N=35 7 (20%) 20 (57%) 5 (14%) 0 (0%) Male Student Harasser, N=35 0 (0%) 26 (74%) 4 (11%) 0 (0%)

Additional Analyses Hypotheses 2 through 5 were also tested using the “no harassment” vignettes

alone in order to further understand the role of power in sexual harassment. The results

for these analyses follow. It is important to reiterate, however, that the “no harassment”

vignettes were not perceived to have contained sexual harassment. For example, many of

the mean values fall at 2 or below, which corresponds to participant responses indicating

that the vignette “clearly did not” portray sexual harassment. Specifically, the female

professor initiator was perceived as having harassed the male student at the level of 3 on

the scale (which corresponds to a “might or might not” response) only about 8% of the

time. The female student initiator was perceived as having harassed the male professor at 78 a level of 3 on the scale less than 3% of the time. The male professor initiator was perceived to have harassed the female student at some level (at levels 3, 4, and 5 on the scale) 14% of the time, while the male student initiator was perceived as having harassed the female professor at a level of 3 on the scale less than 3% of the time.

Hypothesis 2. Table 11 displays the test statistics and means for the test of this hypothesis. When testing Hypothesis 2 including only the “no harassment” vignettes, the multivariate test was non-significant (F(2,139)= 2.35, p>.05), indicating that within the

“no harassment” vignettes, there were no differences between the student and professor initiators on whether either the male or female characters were harassed.

Table 11. Multivariate Analysis of Variance and Means for Type of Initiator, “No Harassment” Vignettes Only. Source F Multivariate tests Initiator Type Wilks’ Lambda (2,139)=.97 2.35 Between subjects StudentProfessor X (SD) X (SD) Initiator Type Male Character Harassed 1.11 1.29 2.18 (0.62) (0.78) Female Character Harassed 1.08 1.33 4.10* (0.60) (0.83) Note. Univariate df=1, 140. Student N= 72; Professor N=70. *p<.05, **p<.01, ***p<.001

79 Hypothesis 3. Analyzing only the “no harassment” conditions in the test of this hypothesis reveals non-significant results (F(25,115)= 1.006, p> .05). Table 12 displays test statistics for this analysis. See Appendix I for means.

Table 12. Multivariate Analysis of Variance for Professor vs. Student Initiators, “No Harassment” Only. Source F Multivariate tests Initiator Type Wilks’ Lambda (25,115)= .82 1.01 Between subjects Initiator Type Female Acted With Intent 2.11 Male Acted With Intent 3.99* Female to Blame 1.35 Male to Blame 0.47 University to Blame 0.35 Male Found Behavior Unwelcome 1.34 Female Found Behavior Unwelcome 1.37 Female Abused Power 0.72 Male Abused Power 1.84 Male Feared Personal Repercussions 0.43 Female Feared Personal Repercussions 6.08* Pretend to be Male Character 0.03 Pretend to be Female Character 0.10 Female Interfered With Environment 0.21 Male Interfered With Environment 0.62 Female Created Intimidating Environment 2.14 Male Created Intimidating Environment 1.19 Female Created Hostile Environment 0.84 Male Created Hostile Environment 1.89 Female Created Offensive Environment 2.14 Male Created Offensive Environment 0.90 Female’s Physical Attractiveness 0.55 Male’s Physical Attractiveness 0.14 Female’s Ethics/Morals 0.00 Male’s Ethics/Morals 1.02 Note. Univariate df=1,137. *p<.05, **p<.01, ***p<.001

80 Hypothesis 4. When only including the “no harassment” conditions in the analysis, there were several initiator gender differences. Table 13 displays the test statistics for this analysis. See Appendix J for means. The multivariate test was significant, F(27,113)=4.327, p<.001. All except four of the variables found to be significant in the main analysis were significant in this analysis with patterns identical to the main analysis, although there were reduced mean differences. The extent to which the male character was harassed, the extent to which the female character acted with intent and was to blame, and the attractiveness of the male character were non-significant in this test of Hypothesis 4. The reduced magnitude of the mean differences indicates that, although initiator gender differences were still present, they were less pronounced, likely due to the fact that the “no harassment” conditions were not perceived to portray harassment.

81 Table 13. Multivariate Analysis of Variance for Male vs. Female Initiators, “No Harassment” Only Source F Multivariate tests Initiator Type Wilks’ Lambda (27,113)=.49 4.33*** Between subjects Initiator Type Male Harassed 0.38 Female Harassed 9.41** Female Acted With Intent 1.28 Male Acted With Intent 4.73* Female to Blame 0.07 Male to Blame 10.01** University to Blame 2.58 Male Found Behavior Unwelcome 13.76*** Female Found Behavior Unwelcome 22.93*** Female Abused Power 6.95** Male Abused Power 13.86*** Male Feared Personal Repercussions 4.83* Female Feared Personal Repercussions 5.75* Pretend to be Male Character 22.48*** Pretend to be Female Character 31.24*** Female Interfered With Environment 19.09*** Male Interfered With Environment 12.15*** Female Created Intimidating Environment 17.13*** Male Created Intimidating Environment 15.29*** Female Created Hostile Environment 4.49* Male Created Hostile Environment 13.14*** Female Created Offensive Environment 9.15** Male Created Offensive Environment 11.44*** Female’s Physical Attractiveness 2.93 Male’s Physical Attractiveness 0.01 Female’s Ethics/Morals 0.45 Male’s Ethics/Morals 0.48 Note. Univariate df=1, 137. Male N= 71; Female N=70. *p<.05, **p<.01, ***p<.001

Hypothesis 5.Analyzing only the “no harassment” vignettes revealed a significant

interaction of initiator gender and initiator position power, F(27,111)= 3.24, p<.001. 82 Table 14 displays the test statistics for this analysis. See Appendix K for means. Only two of the variables which were significant when analyzing only the “harassment” conditions were significant with the same pattern of means; the extent to which the male character abused his power and created an intimidating classroom environment.

Seven additional variables were significant when only analyzing the “no harassment” vignettes. These include: The extent to which the female character was harassed (F(1,137)=7.47, p<.05); the extent to which the male character acted with intent

(F(1,137)= 7.65, p<.05); the extent to which the female character was to blame

(F(1,137)= 3.93 p=.05); the extent to which the female character feared personal repercussions (F(1,137)= 7.16, p<.05); the extent to which the behavior would make the participants uncomfortable if they were the female character (F(1,137)= 5.50, p<.05); the extent to which the female character created an intimidating classroom environment

(F(1,137)= 11.74 p<.05); and the extent to which the male character created a hostile classroom environment (F(1,137)= 6.09 p<.05). Interestingly, many of these variables center on the harassment of the female character by the male character.

The female character was perceived to have been harassed to a greater extent when the initiator was a male professor. Means for the male and female student initiator, as well as the female professor initiator are relatively equivalent. The extent to which the male character acted with intent follows this same pattern, such that the male professor initiator was perceived to have acted with more intent than were the male and female student initiators or the female professor initiator. 83 Female student and professor initiators received relatively equivalent perceptions regarding the extent to which the female character was to blame. The female character was perceived to be to blame to a greater extent when the initiator was a male professor.

The female character was perceived to fear personal repercussions to a greater extent when the vignette contained a male professor initiator. Male and female student initiators, as well as the female professor initiator were equivalent on this variable.

When asked how uncomfortable the participants would be if they were the female character, participants felt that they would be more uncomfortable with the male professor initiator than with the male student initiator. This pattern was reversed, however, for female harassers. The vignette portraying a female student initiator would make the participants more uncomfortable as the female character than would the vignette portraying a female professor initiator. This pattern is not surprising, however, because it is more likely that, as the female character, a participant would feel more uncomfortable in general when a man is initiating behavior than when she is initiating behavior.

The female character was perceived to have created an intimidating classroom environment to a greater extent when the vignette portrayed a female professor initiator.

The remaining 3 vignettes, the female student initiator, and the male student and professor initiators received low, and fairly equivalent, means for the extent to which the female character created an intimidating environment.

Finally, the extent to which the male character was perceived to have created a hostile classroom environment was higher for the vignette portraying a male professor 84 initiator. The remaining 3 vignettes, the female student and professor initiators and the male student initiator were lower, and more equivalent, on this variable.

Table 14. Multivariate Analysis of Variance for Initiator Gender x Power Interaction, “No Harassment Only. Source F Multivariate tests Initiator Gender Wilks’ Lambda (27,111)=.56 3.23*** x Position Power Between subjects Gender x Power Male Harassed 0.41 Female Harassed 7.47** Female Acted With Intent 0.36 Male Acted With Intent 7.65** Female to Blame 3.93* Male to Blame 0.46 University to Blame 0.02 Male Found Behavior Unwelcome 0.35 Female Found Behavior Unwelcome 1.26 Female Abused Power 1.58 Male Abused Power 4.82* Male Feared Personal Repercussions 0.10 Female Feared Personal Repercussions 7.16** Pretend to be Male Character 0.19 Pretend to be Female Character 5.50* Female Interfered With Environment 0.21 Male Interfered With Environment 0.02 Female Created Intimidating 11.74*** Environment Male Created Intimidating Environment 7.56** Female Created Hostile Environment 1.61 Male Created Hostile Environment 6.09* Female Created Offensive Environment 1.71 Male Created Offensive Environment 1.34 Female’s Physical Attractiveness 0.21 Male’s Physical Attractiveness 0.17 Female’s Ethics/Morals 3.24 Male’s Ethics/Morals 0.81 Note. Univariate df=1, 135. Student N= 36; Professor N=34-35. *p<.05, **p<.01, ***p<.001

85 Summary. Although there were several significant differences reported above, it is important to recall that all of the mean values fall below the mid-point on the scale, and that many of them fall below 2, indicating that the vignettes contain little, if any, evidence of harassment. Nevertheless, there was some evidence reflecting the typical script for sexual harassment, such that a male supervisor will be perceived as more harassing than other types of harassers.

Discussion The purpose of this study was to investigate which characteristics of a sexual harassment incident would lead participants to report that harassing behaviors had occurred. Specifically, the goal was to determine the effects of power on evaluations of contrapower sexual harassment. The current study was concerned with determining the effects of position power, harasser and victim gender, participant experience with sexual harassment training, and participant gender on perceptions of sexual harassment.

General Findings Through examination of the participants’ responses, it appears that the manipulations employed in the study were successful. Participants were responding as would be expected for each of the individual conditions. For example, for the vignette portraying a female professor harasser of a male student, the female was perceived as the harasser, and the male was perceived as the victim. Additionally, participants seemed overall to be very cognizant of the position power of the harasser, such that they were aware of circumstances which would allow the professor to abuse his or her power by virtue of his or her supervisory role. Furthermore, examination of the means comparing male and female harassers reveals that, across many variables, male harassers received 86 higher mean ratings (i.e., more negative evaluations) than did the female harassers.

However, these different evaluations were not always statistically significant. Many of the hypotheses presented received either partial or complete support, resulting in several general findings for position power, harasser gender, participant gender, as well as additional findings related to exposure to training and the “no harassment” conditions.

Position Power Consistent with previous research (e.g., Popovich, et al., 1986), in general, the present study indicated that professor harassers were evaluated more negatively than were student harassers. This finding holds for the extent to which harassment occurred, the extent to which each character was to blame, the extent to which the behavior was carried out with intent, the extent to which there was an abuse of power, as well as the extent to which the harasser caused the victim to fear personal repercussions and created an intimidating/offensive environment. There were a few cases, however, in which this pattern did not hold, possibly reflecting an abuse of a non-position-related power. The professor and student harassers were perceived to be equally harassing to a female victim.

Also, participants felt that the male character perceived the behavior to be more unwelcome when coming from a female student harasser than when it came from a female professor harasser. This finding is consistent with Wayne’s (2000) finding that a female subordinate harasser received higher unwelcomeness ratings. Interestingly, in the current study, comparison of the perceptions of female student and professor harassers on this variable revealed that it was largely the female participants who perceived that the female student harasser’s behavior would be more unwelcome. Furthermore, participants felt that the male character feared personal repercussions to a greater extent when the 87 harasser was a female student rather than a female professor, indicating that the male professor was perceived to fear personal repercussions more than was the male student.

An examination of the mean differences between ratings of the female student and professor harassers separately by participant gender revealed no real differences.

Due to the results demonstrating that professor harassers were generally evaluated more negatively than were student harassers, it is likely that participants perceived the professor harassers as abusing their legitimate power, as described by French and Raven

(1959). Because of their supervisory position, professors automatically have legitimate power. Because reward and coercive powers are often related to legitimate power (French and Raven, 1959), it is also possible that participants perceived the possibility of reward and/or coercive power, thus the significant findings for the fear of personal repercussions.

One particular finding reveals evidence for both traditional and contrapower sexual harassment. The male professor and the male student harassers were perceived to be equally harassing to a female victim. On the one hand, the male professor is exhibiting traditional harassment. Furthermore, the male professor appears to possess both legitimate position power as well as societal power stemming from the fact that he is a man. On the other hand, the male student harasser is exhibiting contrapower harassment.

Contrapower harassment is present for cases in which the harasser has less formal power than the victim. In that sense, any student harasser is exhibiting contrapower harassment toward the professor. Because, as a student, he could not have been abusing his position power, participants may have been acknowledging that the male student could operate from another source of power. As noted before, Rospenda, Richman, and Nawyn (1998) state that contrapower sexual harassment also stems from societal power, which is a non- 88 formal source of power. The male student, therefore, may be viewed as a threat by the female victim because he is a man, regardless of the fact that the female is the person with formal power. This finding lends some support for the sociological/societal concept that men are perceived to be more powerful and assertive than are women. Although the male student does not have formal power over his victim, participants may still have perceived him as a threat because of his gender. Ragins and Sundstrom’s (1989) fourth classification of power indicates that sources of power may be additive such that the more sources of power an individual possesses, the more powerful he or she may be. In the case of the male professor, his positional and societal powers may be additive, resulting in an even more powerful presence. The male student does not have quite as much power as the male professor, however, because he has only societal power. In comparison with the female harassers, the male professor harasser will also have more power because, even as a professor, the female harasser only has that one source of power. The female student harasser has the least amount of power because she has neither positional nor societal power. It is also typical for individuals to consider sexual harassment as behaviors that men demonstrate towards women. Therefore, perhaps the student’s gender was more salient to the participants than was his position power. The female character was perceived to be harassed by the male character to an equal extent, regardless of his position power.

Results indicated that the female student also received negative evaluations. As previously mentioned, there is precedent for these findings. Wayne (2000) found in her study that unwelcomeness ratings were higher (i.e., more negative) when the harasser was a subordinate female. The female student was also perceived to have caused the male 89 character to fear personal repercussions. As research by Matchen and DeSouza (2000) has suggested, behaviors from student harassers in an academic setting can have ramifications for their victims. Participants may have perceived that the female student could get her victim into trouble in a variety of ways. For example, the female student’s behaviors may cause other students to feel as though inappropriate behaviors are acceptable in the classroom, thus making the professor’s efforts less effective in maintaining a productive learning environment. Furthermore, the professor’s supervisors may feel that, if a student feels as though she can approach her professor in a sexual way, perhaps the professor is also exhibiting inappropriate behaviors that would indicate he is accepting of sexual advances. Finally, although unlikely in actuality, the professor may fear that if he does not give in to the student’s advances, she may make false accusations of harassment against him.

Dziech and Weiner (1984) discuss possible impacts that the advances of a female student may have on her male professor. The authors state that such behavior can have a significant impact on the professor’s personal and professional lives, although not to the same extent that advances of a professor would have on a student. The effects noted are as follows:

“It may embarrass, annoy, and anger him; it may cause turmoil in both his

private and professional life. But it cannot destroy his self-esteem or endanger is

intellectual self-confidence. Hassled professors do not worry about retaliation or

punitive treatment; they do not fear bad grades or withheld recommendations

from women students. They are not forced to suffer in silence because of fear

of peer disapproval” (pp. 24). 90 Dziech and Weiner do not believe that male professors can be sexually harassed by students. They feel that such an interaction is nothing more than “sexual hassle” because the student does not possess the power necessary to harass. Therefore, unlike the participants in the current study, these authors would not perceive the male character as having a reason to fear personal repercussions stemming from the actions of the female student. They also state that, although there may be some temporary unease, the male professor would not experience the same psychological distress as would a female student in the same situation. The professor has the power in the relationship, and therefore can “control the problem”.

Regardless of the possible personal repercussions that participants may have anticipated for the male character in addition to the examples given above, the current findings indicate evidence for contrapower harassment. Like the case of the male student harasser, the female student harasser could also not have abused her position power.

However, unlike the male student, the female student did not have the advantage of societal power or privilege. As previously mentioned, perhaps the female student was abusing some form of sexual power over her professor. Although some authors (e.g.,

Dziech and Weiner, 1984) would dispute the statement that her behaviors constitute sexual harassment, it is clear that the participants in the current study perceived the male professor to have feared personal repercussions associated with the student’s behaviors.

It is clear from previous literature, such as that by Dziech and Weiner, that there are different perceptions of harassers and their effects on victims based on the gender of the harasser. Whether resulting in sexual harassment or sexual hassle, sexual advances by either gender can be perceived to be inappropriate. It is important to note that gender 91 continues to be an issue in perceptions of sexual harassment, both for harassers and participants.

Gender Harasser gender. It was hypothesized that male harassers would be evaluated more negatively than would female harassers. Results revealed, however, that the harasser was evaluated more negatively, regardless of gender. For instance, when the female character was the harasser, she was evaluated more negatively in the following ways; the male character was perceived to be harassed to a greater extent, she was perceived to be to blame and to have acted with intent to a greater extent, and to have abused her power to a greater extent. Additionally, her behavior was perceived to be more unwelcome, and to have caused the male character to fear personal repercussions.

She was also perceived to have created an intimidating, hostile, and offensive classroom environment, and her behavior would have made the participants uncomfortable, if they were the male character. Finally, the male character was perceived to be more physically attractive when the harasser was a woman.

Similarly, the male character was evaluated more negatively when the male was the harasser. As the harasser, the male character was perceived to have harassed the female character to a greater extent, he was perceived to be to blame to a greater extent, and to have acted with intent to a greater extent. Participants felt that the female character would find the behavior to be unwelcome, that the male harasser abused his power, and caused the female character to fear personal repercussions. When asked to pretend they were the female character, participants also reported that they would have felt uncomfortable with the male character’s behavior. Finally, the male harasser was 92 perceived to have interfered with the female’s classroom environment, and to have created an intimidating, offensive, and hostile environment. There were, however, no significant findings for the attractiveness of the female victim.

Although research (e.g., Popovich, et al., 1995) has demonstrated that a man as the harasser is part of the typical script for a sexual harassment incident, the results of the current study demonstrated that, when women are portrayed as harassers, they will also be perceived as such. Even with men receiving slightly more negative evaluations as the harasser, female harassers were not far behind, and the differences between them were not statistically significant.

It is important to note that, in several cases, the male professor received the most negative evaluations, even in the case of the male professor in the “no harassment” conditions. Although participants did not perceive this character to have harassed the female student, his ratings were still significantly higher than the other “no harassment” conditions. This finding demonstrates the strength of the typical script for sexual harassment. Even when there is no harassment perceived to be present, a supervisory male may still be perceived negatively in his interactions with a subordinate female.

Finally, there were different findings for the attractiveness of the victims of harassment. The male character was perceived to be more physically attractive when the harasser was a woman. Previous research (e.g., Popovich, Gehlauf, Jolton, Everton,

Godinho, Mastrangelo, and Somers, 1996) has found an attractiveness effect in sexual harassment perceptions. It was suggested that there is a myth present where only attractive individuals will be sought out by harassers. It is possible, therefore, that because of the of males as the instigators of harassment, a woman would only 93 harass a man if he were physically attractive. In other words, the participants may have perceived that the male character must have been attractive, or the woman would not have harassed him.

Conversely, there was no finding for the attractiveness of the female character.

Again, because women are typically perceived as the victims of harassment, participants may not have assumed that she must be attractive to be considered a target. Perhaps due to societal power , her gender alone implied that she may be open to sexual advances. As a woman, she may have been perceived as less powerful, thus as an easy target for a man possessing position power, societal power, or both.

Participant gender. As predicted, there were no participant gender differences on perceptions of sexual harassment. The latter finding is consistent with previous literature

(e.g., Rotundo, Nguyen, and Sackett, 2001) which reports that gender differences in perceptions of sexual harassment are typically only found in the less extreme forms of harassment. The current study used behaviors which are generally perceived by both genders as constituting harassment. According to the United States Merit Systems

Protection Board (1995) report, which addressed six broad categories of harassment behaviors: “In virtually every case, the proportion of respondents – both men and women

– who classified the behaviors as sexual harassment rose between 1980 and 1987, and had increased again by 1994” (pp. 5). Some of these increases include men acknowledging more subtle behaviors to be harassment, nearing rates reported by women. These findings indicate that, as previously noted, gender differences in perceptions of harassment have decreased, and may continue to do so in the future. 94 Additional Findings Exposure to sexual harassment training. Contrary to the hypothesized prediction, participants who reported having been exposed to sexual harassment training or education were no more likely to perceive sexual harassment in the vignettes than were participants not having been trained on the subject. One possible reason for this finding could be that the vignettes portrayed such an unambiguous case of sexual harassment that training was not necessary to recognize inappropriate sexual behavior. It is also possible, however, that sexual harassment awareness has been prominent for the vast majority of the participants’ lives. The average age of participants was around 19 years, meaning they were born at least 8 years after the establishment of the EEOC Guidelines on harassment. Perhaps individuals in this age group have been made aware of sexual harassment through means other than formal training or education.

“No harassment” conditions. As previously noted, the data were also examined using only the “no harassment” conditions. While there were some significant findings, nearly all of the mean values fell below 2, which corresponded to a “clearly not” option, indicating that even while the results were statistically significant, they were not necessarily meaningfully significant. However, three findings were slightly above this level, indicating a more uncertain “might or might not” response. Two of these results indicated that the male professor initiator created a hostile environment and caused the female character to fear personal repercussions. These findings are, again, consistent with participants acknowledging the typical script for sexual harassment; a male in a supervisory position harassing his female subordinate. It is important to reiterate, however, that this condition did not receive a full endorsement constituting harassment. 95 The “no harassment” conditions portrayed a student and a professor discussing the student’s grades in the class. While a couple of participants noted in their open-ended response that the characters should not have insisted on being alone during the discussion, the majority of the participants perceived this interaction to portray nothing more than a typical classroom occurrence.

The one remaining significant finding indicated that a female professor initiator created an intimidating environment. This result also has a mean values such that participants did not fully endorse the behavior as a clear case of sexual harassment.

Limitations and Future Research As previously discussed, the behaviors chosen for the vignettes in the current study were used to create an incident of sexual harassment that both male and female participants would be likely to perceive as harassment. Although research (e.g.,

USMSPB, 1995) has shown that gender differences in perceptions of sexual harassment are decreasing, there are still differences present. Because the current study only utilized one form of harassment across all “harassment” conditions, potentially significant and interesting results may have been missed which would have been manifested had a more ambiguous form of harassment had been included. When considering the results that the

“no harassment” conditions were, in some cases, perceived to portray harassment, it is possible that demand characteristics were responsible. Participants were aware that the study was regarding sexual harassment, and may have perceived harassment to be present in an incident where they otherwise may not have perceived harassment. Future research should include additional harassment behaviors, spanning the spectrum from subtle to overt, in order to investigate how contrapower harassment is perceived in different forms. 96 It would also be beneficial to present the information in such a way that it is not so explicit that sexual harassment is the construct of interest.

Furthermore, only two types of harassers were portrayed in the vignettes, supervisory and subordinate. Research (e.g., Popovich, et al., 1986) has shown that while supervisor harassers may be evaluated more negatively, co-worker harassment is more common. This type of harassment was not included in the current study. While it may appear that co-worker harassment would not offer an opportunity to observe contrapower harassment, which was the focus of the current study, it is feasible for contrapower to arise. For instance, between two professors, one professor may have a longer tenure at the university or have obtained a higher level (e.g., full professor versus associate professor).

Similarly, with students, there may be perceived power differences between graduate students and undergraduate students, or even between undergraduate seniors and freshmen. While several interesting results were revealed using only the professor- student dyad, inclusion of finer power distinctions may have also resulted in additional conclusions. Specifically, in addition to professor/student dyads, less clear power relationships such as undergraduate senior/freshman dyads should be included.

Therefore, future research should include co-worker harassment to examine the power distinctions that participants may be making when evaluating cases of sexual harassment.

Although the majority of sexual harassment research has been conducted examining heterosexual interactions, there is also a body of literature that concerns same- sex sexual harassment (SSSH). While perceptions of SSSH are much more common between two men than it is between two women, sexual harassment is not limited to heterosexual encounters. Because the current study’s vignettes only portrayed 97 heterosexual interactions, possible non-gender-related abuses of power were not fully examined. For instance, if a man is sexually harassing one of his male co-workers, both societal and positional abuses of power are debatable, and other avenues must be pursued to explain this behavior. The additive nature of sources of power may play in important role in explaining same-sex harassment. Although some research into SSSH has been conducted (e.g., Stockdale, Berry, Schneider, & Cao, 2004), future research into contrapower harassment should include same-sex interactions to examine the strength of previous conclusions regarding SSSH.

A final limitation of the current study pertains to the sample included in the study.

As previously discussed, the academic setting provides an excellent opportunity for contrapower harassment. Therefore, the choice to use students for data collection provided a chance to analyze the perceptions of the very individuals who have been, or potentially could be, personally involved in an incident similar to that which was portrayed in the vignettes. Similarly, the perceptions of professors are also important for understanding contrapower harassment. Questionnaires were not sent to professors; so valuable perspectives may have been missed. Future research into contrapower harassment should include the viewpoints of professors as a way to search for possible similarities or discrepancies with student evaluations. Because the current study also has implications for non-academic settings, it would be beneficial to replicate the results in an organizational setting.

Implications A better understanding of the many ways in which power can influence how individuals interact with one another has important implications, specifically for sexual 98 harassment education, training, and policy-making. Non-traditional forms of sexual harassment are less often recognized by policy-makers and the public, alike. Sexual harassment education and training need to include the full spectrum of possible behaviors, and students and trainees must be educated about the integral role that power plays in sexual harassment. It is important for individuals to consciously realize that legitimate position power is not the only basis of power that can be abused, resulting in sexual harassment. Organizational and university sexual harassment policies should include descriptions of contrapower and other non-traditional forms of harassment, perhaps even including special grievance procedures to be used due to the unique circumstances surrounding such forms of harassment.

Research into contrapower and other non-traditional forms of harassment, including the current study, help to further understand the role that power plays in sexual harassment. Insights gained via this research are important for the development of future education and training programs on sexual harassment, as well as for the development of new organizational policies, and the revision of old policies, on sexual harassment.

Conclusions Sexual harassment continues to be a problem in organizations, even after nearly

30 years of research and proposed solutions. Sexual harassment reaches beyond the traditional perceptions of a man with power harassing a subordinate woman to encompass many different forms, including contrapower and other non-traditional forms of harassment. Over time, perceptions of harassment have changed, such that both men and women are beginning to acknowledge many of the same behaviors as constituting sexual harassment. It is also not necessary for an individual to have been trained on sexual 99 harassment for him or her to recognize these behaviors when they are present.

Similarly, men and women will both be perceived as harassers: Harassment is no longer a behavior perceived to be perpetrated only by men. Although the typical script for harassment continues to be strongly ingrained in the minds of many individuals, it appears as though people are becoming more aware of other forms of harassment and the multiple bases of power that can be abused as a source of the problem. 100

References Abbey, A. (1987). Misperceptions of friendly behavior as sexual interest: A survey of

naturally occurring incidents. Psychology of Women Quarterly, 11, 173-194.

Bargh, J.A., Raymond, P., Pryor, J.B., & Strack, F. (1995). Attractiveness of the

underling: An automatic power → sex association and its consequences for

sexual harassment and aggression. Journal of Personality and Social

Psychology, 68 (5), 768-781.

Benson, K.A. (1984). Comment on Crocker’s “An analysis of university definitions of

sexual harassment. Signs, 9 (3), 516-519.

Berdahl, J.L., Magley, V.J., & Waldo, C.R. (1996). The sexual harassment of men?

Exploring the concept with theory and data. Psychology of Women Quarterly,

20, 527-547.

Blumenthal, J.A. (1998). The reasonable woman standard: A meta-analytic review of

gender differences in perceptions of sexual harassment. Law and Human

Behavior, 22 (1), 33-57.

Cleveland, J.N. & Kerst, M. E. (1993). Sexual harassment and perceptions of power: an

under-articulated relationship. Journal of Vocational Behavior, 42, 49-67.

Dziech, B.M. & Weiner, L. (1984). The Lecherous Professor: Sexual Harassment

On Campus. Boston: Beacon Press.

DeSouza, E. & Fansler, A.G. (2003). Contrapower sexual harassment: A survey of

students and faculty members. Sex Roles, 48 (11/12), 529-542.

Eagly, A.H. (1983). Gender and social influence: A social psychological analysis.

American Psychologist, 38 (9), 971-981. 101 Equal Employment Opportunity Commission. (1980). Guidelines on discrimination

because of sex (Section 1604.11). Federal Register, 45, 74676-74677.

Equal Employment Opportunity Commission. (2001). Basic facts about sexual

harassment. In L. LaMoncheck & J.P. Sterba (Eds.), Sexual harassment: Issues

and answers (pp. 135). New York: Oxford University Press.

Fitzgerald, L.F., Weitzman, L.F., Gold, Y., & Ormerod, M. (1988). Academic

harassment: Sex and denial in scholarly garb. Psychology of Women Quarterly,

12, 329-340.

French, J. R., & Raven, B. H. (1959). The bases of social power. In D. Cartwright

(Ed.), Studies in social power. Ann Arbor: , Institute of

Social Research.

Gerrity, D.A. (2000). Male university employees’ experiences of sexual harassment-

related behaviors. Psychology of Men & , 1 (2), 140-151.

Gutek, B.A. & Cohen, A.G. (1987). Sex ratios, sex role spillover, and sex at work: A

comparison of men’s and women’s experiences. Human Relations, 40 (2), 97-115.

Gutek, B.A., O’Connor, M.A., Melancon, R., Stockdale, M.S., Geer, T.M., & Done, R.S.

(1999). The utility of the reasonable woman legal standard in hostile

environment sexual harassment cases: A multimethod, multistudy examination.

Psychology, Public Policy, and Law, 5 (3), 596-629.

Ilies, R., Hauserman, N., Schwochau, S., & Stibal, J. (2003). Reported incidence rates of

work-related sexual harassment in the United States: Using meta-analysis to

explain reported rate disparities. Personnel Psychology, 56, 607-631.

Kanter, R.M. (1977). Men and Women of the Corporation. New York: Basic Books, Inc., 102 Publishers.

Levesque, M.J., Nave, C.S., & Lowe, C.A. (2006). Toward an understanding of gender

differences in inferring sexual interest. Psychology of Women Quarterly, 30 (2),

150-158.

Lips, H.M. (1991). Women, Men, and Power. Mountain View, California: Mayfield

Publishing Company.

MacKinnon, C.A. (1979). Sexual Harassment of Working Women. New Haven: Yale

University Press.

Matchen, J. & DeSouza, E. (2000). The sexual harassment of faculty members by

students. Sex Roles, 42 (3/4), 295-306.

Popovich, P.M. (1988). Sexual harassment in organizations. Employee Responsibility and

Rights Journal, 1(4), 273-282.

Popovich, P.M., Gehlauf, D.N., Jolton, J.A., Everton, W.J., Godinho, R.M., Mastrangelo,

P.M., & Somer, J.M. (1996). Physical attractiveness and sexual harassment: Does

every picture tell a story or every story draw a picture? Journal of Applied Social

Psychology, 26, 520-542.

Popovich, P.M, Jolton, J.A., Mastrangelo, P.M., Everton, W.J., & Somers, J.M. (1995).

Sexual harassment scripts: A means to understanding a phenomenon. Sex Roles,

32, 315-335.

Popovich, P.M., Licata, B.J., Nokovich, D., Martelli, T., & Zoloty, S. (1986). Assessing

the incidence and perceptions of sexual harassment behaviors among American

undergraduates. The Journal of Psychology, 120 (4), 387-396.

Pryor, J.B. (1987). Sexual harassment proclivities in men. Sex Roles, 17 (5/6), 260-290. 103 Ragins, B.R. & Sundstrom, E. (1989). Gender and power in organizations: a

longitudinal perspective. Psychological Bulletin, 105 (1), 51-88.

Raven, B.H. (1965). Social influence and power. In I.D. Steiner & M. Fishbein (Eds.),

Current studies in . New York: Hold, Rinehart, Winston.

Reilly, T., Carpenter, S., Dull, V., & Bartlett, K. (1982). The factorial survey: An

approach to defining sexual harassment on campus. Journal of Social Issues,

38 (4), 99-110.

Rospenda, K.M., Richman, J.A., & Nawyn, S.J. (1998). Doing power: The confluence of

gender, race, and class in contrapower sexual harassment. Gender and Society,

12 (1), 40-60.

Rotundo, M., Nguyen, D.H., & Sackett, P.R. (2001). A meta-analytic review of gender

differences in perceptions of sexual harassment. Journal of Applied

Psychology, 86 (5), 914-922. Stockdale 1993

Stockdale, M.S. (1993). The role of sexual misperceptions of women’s friendliness in an

emerging theory of sexual harassment. Journal of Vocational Behavior, 42, 84-

101.

Stockdale, M.S., Berry, C.G., Schneider, R.W., & Cao, F. (2004). Perceptions of the

sexual harassment of men. Psychology of Men & Masculinity, 5(2), 158-167.

Stockdale, M.S., Visio, M., & Batra, L. (1999). The sexual harassment of men: Evidence

for a broader theory of sexual harassment and sex discrimination. Psychology,

Public Policy, and Law, 5 (3), 630-664.

Tangri, S.S. & Hayes, S.M. (1997). Theories of sexual harassment. In William

O’Donohue (Ed.), Sexual harassment: Theory, research, and treatment (pp. 104 112-128). Boston: Allyn and Bacon.

U.S. Merit Systems Protection Board (1994). Sexual harassment of federal workers:

Trends, progress and continuing challenges. Washington, D.C.: United States

Government Printing Office.

Wayne, J.H. (2000). Disentangling the power bases of sexual harassment: Comparing

gender, age, and position power. Journal of Vocational Behavior, 57, 301-325.

Whitley, B.E. (2002). Research in naturalistic settings. In B.E. Whitley, Principles of

Research in Behavioral Science (pp. 336-338). Boston: McGraw Hill.

Wilson, F. & Thompson, P. (2001). Sexual harassment as an exercise of power. Gender

work, and organization, 8 (1), 61-83.

105

Appendix A Equal Employment Opportunity Commission Guidelines on Sexual Harassment

§ 1604.11 Sexual harassment.

(a) Harassment on the basis of sex is a violation of section 703 of title VII. Unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature constitute sexual harassment when (1) submission to such conduct is made either explicitly or implicitly a term or condition of an individual’s employment,

(2) submission to or rejection of such conduct by an individual is used as the basis for employment decisions affecting such individual, or (3) such conduct has the purpose or effect of unreasonably interfering with an individual’s work performance or creating an intimidating, hostile, or offensive working environment.

(b) In determining whether alleged conduct constitutes sexual harassment, the

Commission will look at the record as a whole and at the totality of the circumstances, such as the nature of the sexual advances and the context in which the alleged incidents occurred. The determination of the legality of a particular action will be made from the facts, on a case by case basis.

(c) [Reserved]

(d) With respect to conduct between fellow employees, an employer is responsible for acts of sexual harassment in the workplace where the employer (or its agents or supervisory employees)knows or should have known of the conduct, unless it can show that it took immediate and appropriate corrective action. 106 (e) An employer may also be responsible for the acts of non-employees, with respect to sexual harassment of employees in the workplace, where the employer (or its agents or supervisory employees) knows or should have known of the conduct and fails to take immediate and appropriate corrective action. In reviewing these cases the Commission will consider the extent of the employer’s control and any other legal responsibility which the employer may have with respect to the conduct of such non-employees.

(f) Prevention is the best tool for the elimination of sexual harassment. An employer should take all steps necessary to prevent sexual harassment from occurring, such as affirmatively raising the subject, expressing strong disapproval, developing appropriate sanctions, informing employees of their right to raise and how to raise the issue of harassment under title VII, and developing methods to sensitize all concerned.

(g) Other related practices: Where employment opportunities or benefits are granted because of an individual’s submission to the employer’s sexual advances or requests for sexual favors, the employer may be held liable for unlawful sex discrimination against other persons who were qualified for but denied that employment opportunity or benefit. 107 Appendix B

Vignettes

1. Rebecca, a professor, assigns the seating for the students in her class. She placed William, a student, in the front row. While lecturing, Rebecca continually roams around the classroom, moving up and down each row. One day, Rebecca asked that William remain after class. When they were alone, Rebecca explained to William that if his grades did not improve, he would likely not pass the class.

2. Rebecca, a student, always chose to sit in the front of her professor, William’s, class. Rebecca often seemed to be distracted during class, and William noticed that Rebecca’s would continually drift around the room during the class’s work time. One day, Rebecca stayed after class. When they were alone, Rebecca asked William if her grades would be high enough to pass the class.

3. William, a professor, assigns the seating for the students in his class. He placed Rebecca, a student, in the front row. While lecturing, William continually roams around the classroom, moving up and down each row. One day, William asked that Rebecca remain after class. When they were alone, William explained to Rebecca that if her grades did not improve, she would likely not pass the class.

4. William, a student, always chose to sit in the front of his professor, Rebecca’s, class. William often seemed to be distracted during class, and Rebecca noticed that William’s gaze would continually drift around the room during the class’s work time. One day, William stayed after class. When they were alone, William asked Rebecca if his grades would be high enough to pass the class.

5. Rebecca, a professor, assigns the seating for the students in her class. She placed William, a student, in the front row and continually stands in front of him while lecturing. Rebecca would stand so close to William that she was practically touching him, and William noticed that Rebecca would often stare at him from her desk during the class’s work time. One day, Rebecca asked that William remain after class. When they were alone, Rebecca began making suggestive remarks to William and made it clear that she wanted to have sex with him.

6. Rebecca, a student, always chose to sit in the front of her professor, William’s, class. She would continually try to be physically close to William during class, and William noticed that she would often stare at him from her seat during the class’s work time. One day, Rebecca stayed after class. When they were alone, Rebecca began making suggestive remarks to William and made it clear that she wanted to have sex with him.

7. William, a professor, assigns the seating for the students in his class. He placed Rebecca, a student, in the front row and continually stands in front of her while lecturing. 108 William would stand so close to Rebecca that he was practically touching her, and Rebecca noticed that William would often stare at her from his desk during the class’s work time. One day, William asked that Rebecca remain after class. When they were alone, William began making suggestive remarks to Rebecca and made it clear that he wanted to have sex with her.

8. William, a student, always chose to sit in the front of his professor, Rebecca’s, class. He would continually try to be physically close to Rebecca during class, and Rebecca noticed that he would often stare at her from his seat during the class’s work time. One day, William stayed after class. When they were alone, William began making suggestive remarks to Rebecca and made it clear that he wanted to have sex with her. 109

Appendix C Questionnaire Questions

1. To what extent do you feel that William was sexually harassed?

2. To what extent do you feel that Rebecca was sexually harassed?

3. If you feel that harassment occurred, to what extent do you feel that it was intentional on the part of Rebecca?

4. If you feel that harassment occurred, to what extent do you feel that it was intentional on the part of William?

5. To what extent do you feel that Rebecca is to blame for the incident?

6. To what extent do you feel that William is to blame for the incident?

7. If you feel harassment occurred, to what extent do you feel that the university is to blame for the incident?

8. To what extent do you feel that William found the behavior unwelcome/bothersome?

9. To what extent do you feel that Rebecca found the behavior unwelcome/bothersome?

10. To what extent do you feel that Rebecca abused her power?

11. To what extent do you feel that William abused his power?

12. To what extent do you feel that William might have feared personal repercussions associated with Rebecca’s behavior?

13. To what extent do you feel that Rebecca might have feared personal repercussions associated with William’s behavior?

14. If you were William, how uncomfortable would Rebecca’s behavior make you?

15. If you were Rebecca, how uncomfortable would William’s behavior make you?

16. To what extent do you feel that Rebecca’s behavior interfered with William’s classroom performance?

17. To what extent do you feel that William’s behavior interfered with Rebecca’s classroom performance?

110 18. To what extent do you feel that Rebecca’s behavior created an intimidating classroom environment for William?

19. To what extent do you feel that William’s behavior created an intimidating classroom environment for Rebecca?

20. To what extent do you feel that Rebecca’s behavior created hostile classroom environment for William?

21. To what extent do you feel that William’s behavior created a hostile classroom environment for Rebecca?

22. To what extent do you feel that Rebecca’s behavior created an offensive classroom environment for William?

23. To what extent do you feel that William’s behavior created an offensive classroom environment for Rebecca?

24. How physically attractive do you feel Rebecca is?

25. How physically attractive do you feel William is?

26. How ethical/moral do you feel Rebecca is?

27. How ethical/moral do you feel William is?

28. How old would you estimate Rebecca is? ______years

29. How old would you estimate William is? ______years

30. Please describe briefly how you decided whether this case involved sexual harassment (e.g. your definition of sexual harassment that you used here).

31. Have you been exposed to any sexual harassment education or training?

32. Gender:

33. Age: _____

34. Ethnicity:

35. Major: ______

111 Appendix D Means and Standard Deviations for Manipulations

Variable Vignette Type Mean Std. Dev. Male Character Harassed Male Student Initiator 1.11 0.57 Female Student Initiator 1.11 0.67 Male Professor Initiator 1.20 0.90 Female Professor Initiator 1.37 0.65 Male Student Harasser 1.00 1.08 Female Student Harasser 3.46 0.95 Male Professor Harasser 1.46 1.60 Female Professor Harasser 4.17 0.82 Female Character Harassed Male Student Initiator 1.11 .067 Female Student Initiator 1.06 0.53 Male Professor Initiator 1.66 1.00 Female Professor Initiator 1.00 0.42 Male Student Harasser 4.63 0.60 Female Student Harasser 1.11 0.87 Male Professor Harasser 4.69 .063 Female Professor Harasser 1.00 0.49 Female Acted with Intent Male Student Initiator 0.56 0.69 Female Student Initiator 0.83 1.16 Male Professor Initiator 0.89 0.87 Female Professor Initiator 1.00 1.03 Male Student Harasser 1.17 0.79 Female Student Harasser 4.14 1.19 Male Professor Harasser 1.17 0.57 Female Professor Harasser 4.80 0.41 Male Acted with Intent Male Student Initiator 0.72 0.94 Female Student Initiator .081 1.19 Male Professor Initiator 1.57 1.31 Female Professor Initiator 0.69 0.58 Male Student Harasser 4.74 0.51 Female Student Harasser 1.37 0.84 Male Professor Harasser 4.80 0.41 Female Professor Harasser 1.23 0.69 Female to Blame Male Student Initiator 0.86 0.90 Female Student Initiator 1.19 1.39 Male Professor Initiator 1.49 1.22 Female Professor Initiator 1.06 1.16 Male Student Harasser 1.34 0.64 Female Student Harasser 4.43 0.98 Male Professor Harasser 1.26 0.61 Female Professor Harasser 4.66 0.54 Male to Blame Male Student Initiator 1.56 1.50 112 Female Student Initiator 1.00 1.17 Male Professor Initiator 1.86 1.33 Female Professor Initiator 1.03 1.27 Male Student Harasser 4.66 0.54 Female Student Harasser 1.66 0.91 Male Professor Harasser 4.71 0.52 Female Professor Harasser 1.49 0.78

113

Appendix E Means and Standard Deviations for Hypothesis 3, Part 1

Variable Harasser Type Mean Std. Dev. Female Acted With Intent Student 2.66 1.80 Professor 3.01 1.89 Male Acted With Intent Student 3.06 1.82 Professor 2.99 1.88 Female to Blame Student 2.89 1.76 Professor 2.99 1.80 Male to Blame Student 3.16 1.68 Professor 2.99 1.80 University to Blame Student 1.31 0.73 Professor 1.99 0.92 Male Found Behavior Unwelcome Student 2.10 1.42 Professor 2.29 1.30 Female Found Behavior Unwelcome Student 2.71 1.58 Professor 2.43 1.64 Female Abused Power Student 2.03 1.48 Professor 2.80 1.75 Male Abused Power Student 2.76 1.71 Professor 2.88 1.96 Male Feared Personal Repercussions Student 2.66 1.44 Professor 2.84 1.23 Female Feared Personal Repercussions Student 2.71 1.42 Professor 2.80 1.37 Pretend to be Male Character Student 2.76 1.76 Professor 2.77 1.78 Pretend to be Female Character Student 3.10 1.84 Professor 3.14 1.84 Female Interfered With Environment Student 3.01 1.42 Professor 2.91 1.48 Male Interfered With Environment Student 3.40 1.28 Professor 3.32 1.47 Female Created Intimidating Environment Student 2.59 1.58 Professor 2.84 1.79 Male Created Intimidating Environment Student 2.86 1.57 Professor 2.97 1.80 Female Created Hostile Environment Student 2.04 1.47 Professor 2.41 1.42 Male Created Hostile Environment Student 2.59 1.65 Professor 2.88 1.60 Female Created Offensive Environment Student 2.39 1.51 Professor 2.57 1.54 114 Male Created Offensive Environment Student 2.93 1.50 Professor 3.01 1.70 Female’s Physical Attractiveness Student 2.63 1.79 Professor 2.04 1.80 Male’s Physical Attractiveness Student 2.09 1.82 Professor 1.68 1.60 Female’s Ethics/Morals Student 2.14 1.32 Professor 1.54 1.31 Male’s Ethics/Morals Student 2.03 1.19 Professor 1.75 1.27 Note. Student N=70, Professor N=69.

Means and Standard Deviations for Hypothesis 3, Part 2

Variable Harasser Type Mean Std. Dev. Male Harassed Male Student Harasser (N=35) 4.63 0.60 Female Student Harasser (N=35) 1.11 0.87 Male Professor Harasser (N=34) 4.68 0.64 Female Professor Harasser (N=35) 1.00 0.49 Female Harassed Male Student Harasser (N=35) 1.00 1.08 Female Student Harasser (N=35) 3.46 0.95 Male Professor Harasser (N=34) 1.47 1.62 Female Professor Harasser (N=35) 4.17 0.82 Female Acted With Intent Male Student Harasser (N=35) 1.17 0.79 Female Student Harasser (N=35) 4.14 1.19 Male Professor Harasser (N=34) 1.18 0.58 Female Professor Harasser (N=35) 4.80 0.41 Male Acted With Intent Male Student Harasser (N=35) 4.74 .051 Female Student Harasser (N=35) 1.37 0.84 Male Professor Harasser (N=34) 4.79 0.41 Female Professor Harasser (N=35) 1.23 0.69 Female to Blame Male Student Harasser (N=35) 1.34 0.64 Female Student Harasser (N=35) 4.43 0.98 Male Professor Harasser (N=34) 1.26 0.62 Female Professor Harasser (N=35) 4.66 0.54 Male to Blame Male Student Harasser (N=35) 4.66 0.54 Female Student Harasser (N=35) 1.66 0.91 Male Professor Harasser (N=34) 4.71 0.52 Female Professor Harasser (N=35) 1.49 0.78 University to Blame Male Student Harasser (N=35) 1.37 0.69 Female Student Harasser (N=35) 1.26 0.78 Male Professor Harasser (N=34) 2.06 0.95 Female Professor Harasser (N=35) 1.91 0.89 Male Found Behavior Male Student Harasser (N=35) 1.03 0.62 115 Unwelcome Female Student Harasser (N=35) 3.17 1.15 Male Professor Harasser (N=34) 1.59 1.02 Female Professor Harasser (N=35) 2.97 1.18 Female Found Behavior Male Student Harasser (N=35) 4.09 0.74 Unwelcome Female Student Harasser (N=35) 1.34 0.80 Male Professor Harasser (N=34) 3.82 1.14 Female Professor Harasser (N=35) 1.09 0.56 Female Abused Power Male Student Harasser (N=35) 1.14 0.65 Female Student Harasser (N=35) 2.91 1.56 Male Professor Harasser (N=34) 1.24 0.85 Female Professor Harasser (N=35) 4.31 0.80 Male Abused Power Male Student Harasser (N=35) 3.71 1.64 Female Student Harasser (N=35) 1.80 1.18 Male Professor Harasser (N=34) 4.74 0.51 Female Professor Harasser (N=35) 1.09 0.82 Male Feared Personal Male Student Harasser (N=35) 1.60 1.12 Repercussions Female Student Harasser (N=35) 3.71 0.83 Male Professor Harasser (N=34) 2.12 1.23 Female Professor Harasser (N=35) 3.54 0.74 Female Feared Personal Male Student Harasser (N=35) 3.66 1.14 Repercussions Female Student Harasser (N=35) 1.77 0.97 Male Professor Harasser (N=34) 3.47 1.35 Female Professor Harasser (N=35) 2.14 1.03 Pretend to be Male Male Student Harasser (N=35) 1.31 0.96 Character Female Student Harasser (N=35) 4.20 1.05 Male Professor Harasser (N=34) 1.53 1.35 Female Professor Harasser (N=35) 3.97 1.25 Pretend to be Female Male Student Harasser (N=35) 4.74 0.74 Character Female Student Harasser (N=35) 1.46 0.89 Male Professor Harasser (N=34) 4.79 0.64 Female Professor Harasser (N=35) 1.54 1.01 Female Interfered With Male Student Harasser (N=35) 2.14 1.29 Environment Female Student Harasser (N=35) 3.89 0.96 Male Professor Harasser (N=34) 1.76 1.23 Female Professor Harasser (N=35) 4.03 0.57 Male Interfered With Male Student Harasser (N=35) 4.03 0.71 Environment Female Student Harasser (N=35) 2.77 1.42 116 Male Professor Harasser (N=34) 4.32 0.53 Female Professor Harasser (N=35) 2.34 1.43 Female Created Intimidating Male Student Harasser (N=35) 1.29 0.83 Environment Female Student Harasser (N=35) 3.89 0.96 Male Professor Harasser (N=34) 1.26 0.96 Female Professor Harasser (N=35) 4.37 0.77 Male Created Intimidating Male Student Harasser (N=35) 4.09 0.95 Environment Female Student Harasser (N=35) 1.63 1.00 Male Professor Harasser (N=34) 4.59 0.61 Female Professor Harasser (N=35) 1.40 0.98 Female Created Hostile Male Student Harasser (N=35) 1.23 1.03 Environment Female Student Harasser (N=35) 2.86 1.40 Male Professor Harasser (N=34) 1.29 0.84 Female Professor Harasser (N=35) 3.49 0.95 Male Created Hostile Male Student Harasser (N=35) 3.83 1.20 Environment Female Student Harasser (N=35) 1.34 0.97 Male Professor Harasser (N=34) 4.26 0.71 Female Professor Harasser (N=35) 1.54 0.95 Female Created Offensive Male Student Harasser (N=35) 1.26 0.92 Environment Female Student Harasser (N=35) 3.51 1.07 Male Professor Harasser (N=34) 1.24 0.78 Female Professor Harasser (N=35) 3.86 0.81 Male Created Offensive Male Student Harasser (N=35) 4.03 0.86 Environment Female Student Harasser (N=35) 1.83 1.15 Male Professor Harasser (N=34) 4.53 0.61 Female Professor Harasser (N=35) 1.54 0.95 Female’s Physical Male Student Harasser (N=35) 2.77 1.83 Attractiveness Female Student Harasser (N=35) 2.49 1.76 Male Professor Harasser (N=34) 2.18 1.99 Female Professor Harasser (N=35) 1.91 1.62 Male’s Physical Male Student Harasser (N=35) 1.57 1.67 Attractiveness Female Student Harasser (N=35) 2.60 1.83 Male Professor Harasser (N=34) 0.94 1.13 Female Professor Harasser (N=35) 2.40 1.68 Female’s Ethics/Morals Male Student Harasser (N=35) 2.37 1.70 Female Student Harasser (N=35) 1.91 0.74 Male Professor Harasser (N=34) 1.38 1.79 117 Female Professor Harasser (N=35) 1.69 0.53 Male’s Ethics/Morals Male Student Harasser (N=35) 1.89 0.76 Female Student Harasser (N=35) 2.17 1.50 Male Professor Harasser (N=34) 1.50 0.93 Female Professor Harasser (N=35) 1.89 1.24

118

Appendix F Means and Standard Deviations for Male vs. Female Harassers

Variable Harasser Type Mean Std. Dev. Male Harassed Male 1.23 1.38 Female 3.81 0.95 Female Harassed Male 4.65 0.61 Female 1.06 0.70 Female Acted With Intent Male 1.17 0.69 Female 4.47 0.94 Male Acted With Intent Male 4.77 0.46 Female 1.30 0.77 Female to Blame Male 1.30 0.63 Female 4.54 0.79 Male to Blame Male 4.68 0.53 Female 1.57 0.84 University to Blame Male 1.71 0.89 Female 1.59 0.89 Male Found Behavior Unwelcome Male 1.30 0.88 Female 3.07 1.16 Female Found Behavior Unwelcome Male 3.96 0.96 Female 1.21 0.70 Female Abused Power Male 1.19 0.75 Female 3.61 1.42 Male Abused Power Male 4.22 1.32 Female 1.44 1.07 Male Feared Personal Repercussions Male 1.86 1.19 Female 3.63 0.78 Female Feared Personal Repercussions Male 3.57 1.24 Female 1.96 1.01 Pretend to be Male Character Male 1.42 1.17 Female 4.09 1.15 Pretend to be Female Character Male 4.77 0.69 Female 1.50 0.94 Female Interfered With Environment Male 1.96 1.27 Female 3.96 0.79 Male Interfered With Environment Male 4.17 0.64 Female 2.56 1.43 Female Created Intimidating Environment Male 1.28 0.89 Female 4.13 0.90 Male Created Intimidating Environment Male 4.33 0.83 Female 1.51 0.99 Female Created Hostile Environment Male 1.26 0.93 Female 3.17 1.23 119 Male Created Hostile Environment Male 4.04 1.01 Female 1.44 0.96 Female Created Offensive Environment Male 1.25 0.85 Female 3.69 0.96 Male Created Offensive Environment Male 4.28 0.78 Female 1.68 1.06 Female’s Physical Attractiveness Male 2.48 1.92 Female 2.20 1.70 Male’s Physical Attractiveness Male 1.26 1.45 Female 2.50 1.75 Female’s Ethics/Morals Male 1.88 1.80 Female 1.80 0.65 Male’s Ethics/Morals Male 1.70 0.86 Female 2.09 1.50 Note. Male N=69, Female N=70.

120

Appendix G Means and Standard Deviations for Harasser Gender x Power Interaction

Variable Harasser Harasser Mean Std. Dev. Gender Power Male Harassed Male Student 1.00 1.08 Professor 1.47 1.62 Female Student 3.46 0.95 Professor 4.17 0.82 Female Harassed Male Student 4.63 0.60 Professor 4.68 0.64 Female Student 1.11 0.87 Professor 2.87 0.49 Female Acted With Intent Male Student 1.17 0.79 Professor 1.18 0.58 Female Student 4.14 1.19 Professor 4.80 0.41 Male Acted With Intent Male Student 4.74 0.51 Professor 4.79 0.41 Female Student 1.37 0.84 Professor 1.23 0.69 Female to Blame Male Student 1.34 0.64 Professor 1.26 0.62 Female Student 4.43 0.98 Professor 4.66 0.54 Male to Blame Male Student 4.66 0.54 Professor 4.71 0.52 Female Student 1.66 0.91 Professor 1.49 0.78 University to Blame Male Student 1.37 0.69 Professor 2.06 0.95 Female Student 1.26 0.78 Professor 1.91 .089 Male Found Behavior Unwelcome Male Student 1.03 0.62 Professor 1.59 1.02 Female Student 3.17 1.15 Professor 2.97 1.18 Female Found Behavior Unwelcome Male Student 4.09 0.74 Professor 3.82 1.14 Female Student 1.34 0.80 Professor 1.09 0.56 Female Abused Power Male Student 1.14 0.65 Professor 1.24 0.85 Female Student 2.91 1.56 121 Professor 4.31 0.80 Male Abused Power Male Student 3.71 1.64 Professor 4.74 0.51 Female Student 1.80 1.18 Professor 1.09 0.82 Male Feared Personal Repercussions Male Student 1.60 1.17 Professor 2.12 1.23 Female Student 3.71 0.83 Professor 3.54 0.74 Female Feared Personal Repercussions Male Student 3.66 1.14 Professor 3.47 1.35 Female Student 1.77 0.97 Professor 2.14 1.03 Pretend to be Male Character Male Student 1.31 0.96 Professor 1.53 1.35 Female Student 4.20 1.05 Professor 3.97 1.25 Pretend to be Female Character Male Student 4.74 0.74 Professor 4.79 0.64 Female Student 1.46 0.89 Professor 1.54 1.01 Female Interfered With Environment Male Student 2.14 1.29 Professor 1.76 1.23 Female Student 3.89 0.96 Professor 4.03 0.57 Male Interfered With Environment Male Student 4.033 0.71 Professor 4.32 0.53 Female Student 2.77 1.42 Professor 2.34 1.43 Female Created Intimidating Male Student 1.29 0.83 Environment Professor 1.26 0.96 Female Student 3.89 0.96 Professor 4.37 0.77 Male Created Intimidating Male Student 4.09 0.95 Environment Professor 4.59 0.61 Female Student 1.63 1.00 Professor 1.40 0.98 Female Created Hostile Environment Male Student 1.23 1.03 Professor 1.29 0.84 Female Student 2.86 1.40 Professor 3.49 0.95 Male Created Hostile Environment Male Student 3.83 1.20 Professor 4.26 0.71 122 Female Student 1.34 0.97 Professor 1.54 0.95 Female Created Offensive Male Student 1.26 0.92 Environment Professor 1.24 0.78 Female Student 3.51 1.07 Professor 3.86 0.81 Male Created Offensive Environment Male Student 4.03 0.86 Professor 4.53 0.61 Female Student 1.83 1.15 Professor 1.54 0.95 Female’s Physical Attractiveness Male Student 2.77 1.83 Professor 2.18 1.99 Female Student 2.49 1.76 Professor 1.91 1.62 Male’s Physical Attractiveness Male Student 1.57 1.67 Professor 0.94 1.13 Female Student 2.60 1.83 Professor 2.40 1.68 Female’s Ethics/Morals Male Student 2.37 1.70 Professor 1.38 1.79 Female Student 1.91 0.74 Professor 1.69 0.53 Male’s Ethics/Morals Male Student 1.89 0.76 Professor 1.50 0.93 Female Student 2.17 1.50 Professor 2.00 1.51 Note. Student N=35, Professor N=34. 123

Appendix H Means and Standard Deviations for Participant Gender

Variable Harasser Gender Harasser Power Mean Std. Dev. Male Student Initiator Male Harassed Male (N=11) 0.91 0.70 Female (N=23) 1.22 0.52 Female Harassed Male (N=11) 0.91 0.70 Female (N=23) 1.22 0.67 Female Student Initiator Male Harassed Male (N=15) 1.13 0.52 Female (N=21) 1.10 0.77 Female Harassed Male (N=15) 1.07 0.46 Female (N=21) 1.05 0.59 Male Professor Initiator Male Harassed Male (N=13) 1.08 0.49 Female (N=21) 1.29 1.10 Female Harassed Male (N=13) 1.23 0.44 Female (N=21) 1.95 1.16 Female Professor Initiator Male Harassed Male (N=17) 1.29 0.59 Female (N=18) 1.44 0.70 Female Harassed Male (N=17) 1.12 0.33 Female (N=18) 0.89 0.47 Male Student Harasser Male Harassed Male (N=15) 0.80 0.56 Female (N=20) 1.15 1.35 Female Harassed Male (N=15) 4.53 0.74 Female (N=20) 4.70 0.47 Female Student Harasser Male Harassed Male (N=17) 3.18 0.88 Female (N=18) 3.72 0.96 Female Harassed Male (N=17) 1.06 0.56 Female (N=18) 1.17 1.10 Male Professor Harasser Male Harassed Male (N=15) 1.93 1.75 Female (N=19) 1.11 1.45 Female Harassed Male (N=15) 4.47 0.83 Female (N=19) 4.84 0.37 Female Professor Harasser Male Harassed Male (N=18) 3.94 0.73 Female (N=17) 4.41 0.87 Female Harassed Male (N=18) 1.00 0.34 Female (N=17) 1.00 0.61

124

Appendix I Means and Standard Deviations for Professor vs. Student Initiators

Variable Harasser Type Mean Std. Dev. Female Acted With Intent Student 0.69 0.96 Professor 0.93 0.94 Male Acted With Intent Student 0.76 1.07 Professor 1.13 1.11 Female to Blame Student 1.03 1.17 Professor 1.26 1.21 Male to Blame Student 1.28 1.37 Professor 1.43 1.37 University to Blame Student 0.75 0.98 Professor 0.84 0.83 Male Found Behavior Unwelcome Student 1.49 1.06 Professor 1.71 1.24 Female Found Behavior Unwelcome Student 1.67 1.19 Professor 1.91 1.31 Female Abused Power Student 1.14 0.92 Professor 1.28 0.98 Male Abused Power Student 1.31 0.93 Professor 1.55 1.21 Male Feared Personal Repercussions Student 1.60 1.15 Professor 1.72 1.16 Female Feared Personal Repercussions Student 1.49 1.01 Professor 1.99 1.38 Pretend to be Male Character Student 1.44 0.96 Professor 1.42 0.85 Pretend to be Female Character Student 1.74 1.09 Professor 1.80 1.22 Female Interfered With Environment Student 1.72 1.21 Professor 1.81 1.12 Male Interfered With Environment Student 2.07 1.13 Professor 1.91 1.23 Female Created Intimidating Environment Student 1.61 1.00 Professor 1.88 1.21 Male Created Intimidating Environment Student 1.82 1.09 Professor 2.04 1.34 Female Created Hostile Environment Student 1.35 0.91 Professor 1.49 0.98 Male Created Hostile Environment Student 1.50 0.92 Professor 1.75 1.25 Female Created Offensive Environment Student 1.42 0.88 Professor 1.65 1.03 125 Male Created Offensive Environment Student 1.58 1.00 Professor 1.75 1.13 Female’s Physical Attractiveness Student 1.76 1.75 Professor 1.55 1.66 Male’s Physical Attractiveness Student 1.46 1.56 Professor 1.36 1.51 Female’s Ethics/Morals Student 2.19 1.56 Professor 2.20 1.49 Male’s Ethics/Morals Student 2.25 1.42 Professor 2.00 1.52 Note. Student N=72, Professor N=69. 126

Appendix J Means and Standard Deviations for Male vs. Female Initiators, “No Harassment” Only

Variable Harasser Mean Std. Dev. Type Male Harassed Male 1.15 0.75 Female 1.23 0.66 Female Harassed Male 1.38 0.88 Female 1.01 0.47 Female Acted With Intent Male 0.72 0.80 Female 0.90 1.09 Male Acted With Intent Male 1.14 1.21 Female 0.74 0.94 Female to Blame Male 1.17 1.11 Female 1.11 1.28 Male to Blame Male 1.70 1.42 Female 1.00 1.22 University to Blame Male 0.92 0.97 Female 0.67 0.83 Male Found Behavior Unwelcome Male 1.25 1.00 Female 1.94 1.20 Female Found Behavior Unwelcome Male 2.25 1.35 Female 1.31 0.94 Female Abused Power Male 1.00 0.72 Female 1.41 1.11 Male Abused Power Male 1.75 1.26 Female 1.10 0.73 Male Feared Personal Repercussions Male 1.45 1.08 Female 1.87 1.19 Female Feared Personal Repercussions Male 1.97 1.34 Female 1.49 1.05 Pretend to be Male Character Male 1.10 0.64 Female 1.77 1.01 Pretend to be Female Character Male 2.25 1.23 Female 1.27 0.82 Female Interfered With Environment Male 1.37 0.99 Female 2.17 1.19 Male Interfered With Environment Male 2.32 1.18 Female 1.66 1.09 Female Created Intimidating Environment Male 1.38 0.88 Female 2.11 1.20 Male Created Intimidating Environment Male 2.31 1.29 Female 1.54 1.02 Female Created Hostile Environment Male 1.25 0.86 127 Female 1.59 1.00 Male Created Hostile Environment Male 1.94 1.19 Female 1.30 0.89 Female Created Offensive Environment Male 1.30 0.78 Female 1.77 1.07 Male Created Offensive Environment Male 1.96 1.16 Female 1.37 0.87 Female’s Physical Attractiveness Male 1.90 1.79 Female 1.41 1.58 Male’s Physical Attractiveness Male 1.42 1.44 Female 1.40 1.64 Female’s Ethics/Morals Male 2.11 1.61 Female 2.29 1.44 Male’s Ethics/Morals Male 2.04 1.49 Female 2.21 1.46 Note. Male N=71, Female N=70. 128

Appendix K Means and Standard Deviations for Initiator x Position Power Interaction, “No Harassment” Only

Variable Initiator Initiator Mean Std. Dev. Gender Power Male Harassed Male Student 1.11 0.57 Professor 1.20 0.90 Female Student 1.11 0.67 Professor 1.35 0.65 Female Harassed Male Student 1.11 .067 Professor 1.66 1.00 Female Student 1.06 0.53 Professor 0.97 0.39 Female Acted With Intent Male Student 0.56 .069 Professor 0.89 0.87 Female Student 0.83 1.16 Professor 0.97 1.03 Male Acted With Intent Male Student 0.72 0.94 Professor 1.57 1.31 Female Student 0.81 1.19 Professor 0.68 0.59 Female to Blame Male Student 0.86 0.90 Professor 1.49 1.22 Female Student 1.19 1.39 Professor 1.03 1.17 Male to Blame Male Student 1.56 1.50 Professor 1.86 1.33 Female Student 1.00 1.17 Professor 1.00 1.28 University to Blame Male Student 0.86 1.02 Professor 0.97 0.92 Female Student 0.64 0.93 Professor 0.71 0.72 Male Found Behavior Unwelcome Male Student 1.19 0.98 Professor 1.31 1.02 Female Student 1.78 1.07 Professor 2.12 1.32 Female Found Behavior Unwelcome Male Student 2.03 1.30 Professor 1.49 1.38 Female Student 1.31 0.95 Professor 1.32 0.94 Female Abused Power Male Student 1.03 0.70 Professor 0.97 0.75 129 Female Student 1.25 1.11 Professor 1.59 1.10 Male Abused Power Male Student 1.44 1.05 Professor 2.06 1.39 Female Student 1.17 0.77 Professor 1.03 0.67 Male Feared Personal Repercussions Male Student 1.42 1.11 Professor 1.49 1.07 Female Student 1.78 1.17 Professor 1.97 1.22 Female Feared Personal Repercussions Male Student 1.47 0.94 Professor 2.49 1.50 Female Student 1.50 1.08 Professor 1.47 1.02 Pretend to be Male Character Male Student 1.14 0.68 Professor 1.06 0.59 Female Student 1.75 1.11 Professor 1.79 0.91 Pretend to be Female Character Male Student 2.03 1.11 Professor 1.49 1.31 Female Student 1.44 1.00 Professor 1.09 0.51 Female Interfered With Environment Male Student 1.36 1.10 Professor 1.37 0.88 Female Student 2.08 1.23 Professor 1.81 1.12 Male Interfered With Environment Male Student 2.42 1.05 Professor 2.23 1.31 Female Student 1.72 1.11 Professor 1.59 1.08 Female Created Intimidating Male Student 1.53 0.94 Environment Professor 1.23 0.81 Female Student 1.69 0.06 Professor 2.56 1.19 Male Created Intimidating Male Student 1.94 1.04 Environment Professor 2.69 1.43 Female Student 1.69 1.14 Professor 1.38 0.85 Female Created Hostile Environment Male Student 1.28 0.91 Professor 1.23 0.81 Female Student 1.42 0.91 Professor 1.76 1.07 Male Created Hostile Environment Male Student 1.61 0.90 130 Professor 2.29 1.36 Female Student 1.39 0.93 Professor 1.21 0.84 Female Created Offensive Male Student 1.28 0.74 Environment Professor 1.31 0.83 Female Student 1.56 1.00 Professor 2.00 1.10 Male Created Offensive Environment Male Student 1.78 0.99 Professor 2.14 1.31 Female Student 1.39 0.99 Professor 1.35 0.73 Female’s Physical Attractiveness Male Student 1.94 1.84 Professor 1.86 1.77 Female Student 1.58 1.66 Professor 1.24 1.50 Male’s Physical Attractiveness Male Student 1.42 1.50 Professor 1.43 1.40 Female Student 1.50 1.65 Professor 1.29 1.64 Female’s Ethics/Morals Male Student 2.33 1.67 Professor 1.89 1.53 Female Student 2.06 1.45 Professor 2.53 1.40 Male’s Ethics/Morals Male Student 2.06 1.33 Professor 2.03 1.65 Female Student 2.44 1.50 Professor 1.97 1.40 Note. Student N=36, Professor N=34-35.