Sexual Harassment of Professional Women: 

The Effects of Physical Appearance on Perceptions of Sex Role Beliefs


Megan J. Moore  and  Connie T. Wolfe

Middle Tennessee State University and Hanover College


Poster presented at the 3rd annual meeting of

the Society for Social and Personality Psychologists (Feb., 2002, Savannah, GA).

This poster is based on a senior thesis completed by Megan J. Moore

at Hanover College during the 2001-2002 school year.


Please address all correspondence to:

Megan Moore at



The present research explored the influence of females’ physical appearance on males’ perceptions of their sex role beliefs and males’ willingness to endorse sexually harassing behaviors in scenarios. Study 1 found support for the idea that a woman’s physical appearance does cue sex role beliefs.  Study 2 found some support for the notion that women who appear to be more egalitarian may be more likely to be the recipients of sexually harassing behaviors. Study 2 also found that men who scored higher on a sexism scale were more likely to say they would engage in sexually harassing or negative behaviors toward professional women than men scoring lower on a sexism scale.  Results imply that women who dress in an “egalitarian way” in the workplace may be more susceptible to sexual harassment than “traditionally dressed” women.  Results also point to the need for researchers to address the issue of how messages about a woman’s sex role beliefs are conveyed.



          Sexual harassment can take the form of “quid pro quo” extortion of sexual favors, unwanted sexual attention, or gender harassment (United States Merit Systems Protection Board, USMSPB, 1995).  Sexual harassment tends to strike highly educated women in traditionally male-dominated professions. For example, according to the USMSPB (1995), federally employed women who were “trainees, blue-collar workers, office workers, service workers” were less likely than professionals, administrators and managers to have experienced sexual harassment in the previous two years.  Rosenberg et al. (1993) found that approximately two-thirds of the 220 female lawyers in their sample reported being addressed as “honey” or “dear” and being the target of remarks emphasizing gender and sexuality (e.g., “nice to have a pretty face”) in professional situations.  Dall and Maass (1999) provide experimental evidence minimizing the concern that professional women simply report sexual harassment more than nonprofessional women. They found that a woman described as egalitarian (an accountant) was more likely to be sexually harassed than a woman described as traditional (an elementary school teacher).

          Why are professional women in nontraditional careers more likely to be sexually harassed?  Some researchers (Dall & Maass, 1999; Rosenberg et al., 1993) suggest that some male colleagues use sexual harassment as a means of preserving their dominant status which professional women threaten. Gutek and Marasch (1982) have termed the phenomenon that people expect women in the workplace to behave according to traditional sex role expectations instead of masculine work roles as “sex role spillover.”  Women’s work roles are expected to be congruent with their sex roles in that they are to be more submissive, nurturing, sympathetic and loyal than men in the same work roles. Gutek and Morasch argue that sexual harassment operates according to this theory particularly when gender is salient. When the sex ratio is skewed (as with women in nontraditional occupations) then “spillover from the sex role of that sex to the work role occurs” (p. 63).

Perhaps women in nontraditional jobs are also projecting nontraditional sex role beliefs, thus further threatening men's dominance.  Women who are viewed as being non-traditional in some way are judged more harshly by their [sexist] male colleagues (Dall & Maass, 1999; Etaugh, 1973, Haddock & Zanna, 1994; Kaley, 1971; Swim & Cohen, 1997). 

Could women’s sex role beliefs be conveyed simply by physical appearance? Fiske et al. (1991) provide anecdotal evidence that women may be received more positively in the workplace by sexually harassing men if they dress more traditionally. In Price Waterhouse v. Hopkins, Ann Hopkins claimed that accounting partners described her as a “macho” woman and one partner advised her to “dress more femininely, wear make-up, have her hair styled, and wear jewelry” (p. 1050).

          There is a lack of research where experimenters have elicited the impression that a woman violates or complies with sex role expectations and, in turn, whether sexual harassment is a response to that impression.  Rather, researchers have already done the work for participants, so to speak, by recording the likelihood of sexually harassing women labeled as “feminist,” “egalitarian,” or “traditional” or women fitting written descriptions of those labels.

       The present studies were designed to address these two questions:

(1) Do women convey a certain sex role attitude simply by their appearance?

(2) Are women who appear to be more egalitarian in their sex role beliefs more likely to be targets for sexually harassing behaviors?


The first question is addressed in Study 1 and the second in Study 2, a scenario study.



Design & Method

          Research literature indirectly suggests that women who are wearing dark colored, masculine clothing (e.g., suit) and who have short hair may be more likely to be perceived as egalitarian or nontraditional (cf. Deaux & Lewis, 1984; Morton, 1964; Terry & Krantz, 1993). Similarly, women who have longer hair and are dressed in lighter, more feminine (e.g., sweater & dress) clothing may be perceived as having more traditional sex role attitudes.

          The present study tested this hypothesis by creating computer generated images of women matching the above descriptions.[1] The women depicted in Figures 1 & 2 were hypothesized to be viewed as egalitarian and the women depicted in Figures 3 & 4 were hypothesized to be traditional. (Two other pictures that will not be discussed here were also used in this study.)

          Seventy-three undergraduates (33 males and 40 females, mean age = 20.4 years) volunteered their time to participate. Each participant received one picture and was asked to respond to the Modern Sexism (MS) scale (Swim, Aikin, Hall & Hunter, 1995) with the following directions: “Please respond to each statement the way you feel the woman pictured would respond.”  Responses were given on a 1-7 Likert scale with lower numbers indicating more egalitarian beliefs. For example, they were asked to indicate the extent to which the woman in the picture would agree or disagree with the statement:  “Women often miss out on good jobs due to sexual discrimination.”

Results & Discussion

          A one-way analysis of variance indicated a significant effect of picture on MS scores, F(5,67) = 4.90, p < .05.  The pattern of means supported our hypothesis that Figures 1(M = 2.56) & 2 (M = 3.16) would be rated as most egalitarian and Figures 3 (M = 3.65) & 4 (M = 3.84) would be rated as most traditional.[2] Thus, while this is not as strong of evidence as we may have liked, these findings do support the notion that women may convey information about their sex role beliefs simply via their outward appearance.



Design & Method

          Study 2 was designed to test the hypothesis, derived from sex role spillover theory (Gutek & Morasch, 1982),  that men will be more likely to sexually harass women who appear to hold egalitarian sex role beliefs as compared to women who appear to hold traditional sex role beliefs.

          Forty-five males (mean age = 19.77 years) volunteered their time. Five of those participants received extra credit for their participation.

          The men filled out the Modern Sexism Scale (Swim et al., 1995) and the Ambivalent Sexism Inventory (ASI; Glick & Fisk, 1997).  Additionally, the men were given one of the 4 pictures of women (Figures 1-4) and presented with a packet of five hypothetical scenarios (see Appendix for an example).  These scenarios were created and adapted from the Likelihood to Sexually Harass (LSH) Questionnaire (Pryor, 1987).  Male participants were asked to imagine themselves involved in each scenario with the woman pictured as the female professional (i.e., associate in a law firm, author, surgeon, biology professor, or chemistry lab assistant). Following each scenario, five alternative courses of action were listed, one of which was an explicitly sexually harassing behavior.  Participants were asked to indicate the likelihood of performing each of the behaviors on a 1-8 Likert type scale with higher numbers indicating a likelihood to do the action. The scenarios were presented in random order and the sexism scales were randomly presented before or after the scenarios.  A manipulation check on this sample’s ratings of the sex role beliefs of Figures 1-4 was presented last to participants.


          While there were no statistically significant differences by picture for the manipulation check (F < 1), the means were in a direction similar to that of Study 1 and did not contradict our assumptions that Figure 1 would be viewed as the most egalitarian female, followed by Figure 2, 3 and 4. 

Not surprisingly, the males in the sample self-reported as low sexists, both on the MS (M = 3.83, SD = .79) and the overall ASI (M = 2.55, SD = .49).  Additionally, male participants were generally unlikely to sexually harass or negatively treat any of the women when looking at the combined scores of all 5 courses of action (reverse coded where appropriate) across all 5 scenarios (M = 1.89, SD = 1.18).

Across all analyses conducted, several findings were of note and the most interesting will be summarized here. There were no significant findings when collapsing across scenario. However, we also examined each behavior for each scenario separately using 2 (hi/low sexism) X 4 (woman pictured) ANOVAs.  In examining responses to the scenario involving the woman as a biology professor (see Appendix),  there was a main effect of picture, F (3,36) = 4.23, p < .05.  An SNK post hoc test indicated that the second most egalitarian woman (Figure 2) was more likely than any of the other three to be sexually harassed (see Table 1).  A median split on modern sexism scores yielded no significant main effect, nor was there an interaction between picture and MS. This same pattern of results was found for the ASI overall as well as both its benevolent sexism subscale and hostile sexism subscale. No significant findings of note were found for the sexual harassment behavior for the other 4 scenarios. 

In examining the other possible (non-sexually harassing) behaviors for each scenario, we found no significant effects for the surgeon or author scenarios, but a few significant findings for the lab assistant, lawyer and biology professor scenarios.  In each case, the finding was such that high sexists (as scored by one or more of the scales) would be more likely to engage in a negative behavior toward the woman than low sexists. For example, hostile sexists were more likely (M = 4.43) than low hostile sexists (M = 3.05) to offer the female lawyer in one scenario a promotion, but only with the condition of starting her at a lower salary than a male counterpart; F(1, 37) = 6.76, p < .05. Hostile sexists were also more likely (M = 3.08) to take everything the biology professor said as a personal attack than were low hostile sexists (M = 2.11).


          Results indicate that participants were more willing to engage in a sexually harassing behavior toward a biology professor when she was illustrated by picture of a woman rated as more egalitarian rather than a more traditionally rated picture. This significant main effect provides some partial support for the hypothesis that women who simply appear egalitarian will be more likely to be sexually harassed. One reason this particular scenario may have yielded results while the others did not is that the biology professor scenario may have seemed more real to the male college students involved and, thus, elicited slightly more honest responding. Unlike imagining discussing their grade with a professor, the participants may have had difficulty imagining themselves in the role of a partner of a law firm, an editor of an author’s book, or a surgeon’s co-worker. As is the case with all scenario studies, self-report bias was likely a big problem here. The value of making the scenario “real” to participants should not be underestimated. 

          Similarly, the non-sexually harassing items may have elicited fewer demand characteristics and, thus, less socially desirable responding from male participants who scored higher on the sexism scales.  It is puzzling, however, that the negative behavior endorsed as “likely” by high sexist males was not directed more specifically toward women pictured as egalitarian.



          So why are professional women more likely to be at risk for sexual harassment than women in more traditionally female occupations?  Perceptions of a woman’s own sex role beliefs certainly play a role in fueling sex role spillover effects. The present two studies raise the interesting possibility that men may not need very much information to feel they know something about a woman’s sex role beliefs. Indeed, men may simply look at a woman’s physical appearance and feel they know whether that woman is willing to conform or not to traditional sex role attitudes and behaviors. With respect to actual sexual harassment, men’s own sex role beliefs predicted some negative and/or sexually harassing behaviors, and perceptions of the woman’s beliefs independently influenced the likelihood of sexual harassment in one scenario.  It is puzzling, however, that these two factors did not interact.  The reported study consisted of a small sample from a very homogenous and small liberal arts college.  Much stronger results would likely be found if this study were repeated in a sample of professional men, perhaps also with stimulus materials targeted toward actual experiences these professional men may find themselves in. The present study does point to the necessity for researchers to attend carefully to how certain characteristics of female targets are portrayed in their research.





Imagine that the woman pictured is your biology professor.  She is the only female professor in the biology department.  You are making an average grade in her class. You are sitting a few feet apart in her office discussing how you can improve your grade.  How likely are you to do the following?


a) Tell her you appreciate her helping you.

b)  Ask her if she has a boyfriend and, if not, would she like one?(*)

c)  Take everything she says as a personal attack.

d)  Scoot your chair back a few inches.

e)  Tell her you do not like her approach to teaching.


(*) The explicitly sexually harassing item.





Means for main effect of pictures on likelihood to sexually harass biology professor


Picture                                                       Likelihood to Sexually Harass

Figure 1 (Egalitarian)                                         1.42

Figure 2 (Egalitarian)                                         3.09*

Figure 3 (Traditional)                                        1.63 

Figure 4 (Traditional)                                        1.50








Dall, A. & Maass, A. (1999).  Studying sexual harassment in the laboratory: Are egalitarian women at higher risk?  Sex Roles, 41, 681-704.

Deaux, K. & Lewis, L.L. (1984).  Structure of gender stereotypes: Interrelationships among compenents and gender label.  Journal of Personality and Social Psychology, 46, 991-1004.

Etaugh, C.F. (1973).  Attitudes of professionals toward the married professional woman.  Psychological Reports, 32, 775-780.

Fiske, S.T., Bersoff, D.N., Borgida, E., Deaux, K. (1991). Social science research on trial: Use of sex stereotyping research in Price Waterhouse v. Hopkins.  American Psychologist, 46, 1049-1060.

Gutek, B.A. & Morasch, B. (1982).  Sex ratios, sex-role spillover, and sexual harassment of women at work.  Journal of Social Issues, 38, 55-74.

Haddock, G. & Zanna, M.P. (1994).  Preferring “housewives” to “feminists”: Categorization and the favorability of attitudes toward women.  Psychology of Women Quarterly, 18, 25-52.

Kaley, M.M. (1971).  Attitudes toward the dual role of the married professional woman. American Psychologist, 26, 301-306.

Morton, G.M. (1964).  The art of costume and personal appearance.  New York:  University of Nebraska Foundation.

Rosenberg, J. Perlstandt, H. & Phillips, W.R. (1993).  Now that we are here: Discrimination, disparagement and harassment at work and the experience of women lawyers. Gender and Society, 7, 415 – 433.

Swim, J.K., Aikin, K.J., Hall, W.S. & Hunter, B.A. (1995). Sexism and racism: Old-fashioned and modern prejudices.  Journal of Personality and Social Psychology, 68, 199-214.

Swim, J.K. & Cohen, L.L. (1997).  Overt, covert and subtle sexism: A comparison between the attitudes toward women and modern sexism scales.  Psychology of Women Quarterly, 21, 103-118.

Terry, R. L. & Krantz, J.H. (1993).  Dimensions of trait attributions associated with eyeglasses, men’s facial hair, and women’s length of hair.  Journal of Applied Social Psychology, 23, 1757-1769.

United States Merit Systems Protection Board (1995). Sexual harassment in the federal workplace:  Trends, progress and continuing challenges.  Washington, D.C.: United States Government Printing Office.

[1] “Cosmopolitan Fashion Makeover Deluxe” and “Cosmospolitan Virtual Makeover Deluxe 2” were used to create the images.

[2] A Student-Newman Keuls post hoc indicated that Figure 1 was rated significantly  more egalitarian than the other pictures in the study. A least significant difference post hoc test indicated that Figure 2 and Figure 4 were rated differently from each other.