Sex Roles, VoL 37, Nos. 9/10, 1997
Sexual Harassment: Factors Affecting Attitudes and Perceptions Danielle Foulis and Marita P. McCabe 1
Deakin University
The current study investigated the effects of gender, gender role, gender role stereotypes, age, occupation, and experience of sexual harassment on both attitudes to and perceptions of sexual harassment. The effects of these variables were also investigated in relation to experience of sexual harassment. Five questionnaires were administered to a sample of 48 high school students, 73 university students, and 75 workers (123 females, 73 males). The majority of respondents were of Anglo-Saxon descent. Gender role, gender role stereotypes, experience of sexual harassment, and perceptions of sexual harassment predicted attitudes to sexual harassment. Gender differences were also found in sexual harassment attitudes. Perceptions of sexual harassment were predicted by attitudes to sexual harassment, and gender role, with no gender differences being exhibited. Occupation, attitudes to sexual harassment and gender role were found to predict experience of sexual harassment. Age was also found to be strongly related to this variable. The results are interpreted within the framework of social learning theory and script theory. Limitations of the study and recommendations for future research are made.
Although there is an increasing a m o u n t of literature on sexual harassment, there is no one universally accepted definition. As aptly put by Roscoe, Strouse, and Goodwin, It is widely argued that problems in understanding and addressing sexual harassment stem from the lack of a clear, concise, universally accepted definition of sexual harassment. (1994, p. 516)
1To whom correspondence should be addressed at School of Psychology, Deakin University, 221 Burwood Highway, Burwood, Victoria, Australia 3125. 773 0360-0025/97/1100-0773512.50/09 1997 Plenum Publishing Corporation
774
Foulis and McCabe
The lack of agreement among researchers on what constitutes sexual harassment is reflective of the lack of agreement from one person to another in the general population. What one person defines as sexual harassment, another person may not, and thus it is not surprising that a vast amount of the research conducted in this area has concentrated on determining what exactly affects a person's perception of sexual harassment. While extreme cases of sexual harassment are often agreed upon and defined as such, it is the more moderate forms of sexually harassing behavior that produce differences in attitudes and conceptualisations (Rubin & Borgers, 1990). For the purposes of this study, the definition of sexual harassment was taken from guidelines provided by the Australian Human Rights and Equal Opportunity Commission. Sexual harassment is any unwanted or uninvited sexual behavior which is offensive, embarrassing, intimidating or humiliating. It has nothing to do with mutual attraction or friendship. ("The Sex Discrimination Act 1984: Sexual Harassment--Knowing Your Rights")
While this definition is also subjective, in that it is essentially the recipient of the behavior who defines the incident as sexual harassment in accordance with how they perceive it, it is the behaviors that are listed with this definition that will be used in the present study: touching, patting, or fondling; staring or leering; sexually suggestive comments and jokes; persistent invitations or requests for sex; showing lewd pictures, posters, or cartoons; and unnecessary familiarity, such as deliberately brushing up against someone. Lott, Reilly, and Howard (1982) found that 29.4% of female respondents reported being sexually assaulted at least once as opposed to 5.3% of male respondents. The term "sexual assault" in this study included behaviors defined as sexual harassment in the current study. In the same survey, only 30% of women reported that they had never been sexually insulted by a man on campus. In a study by Reilly, Lott, Caldwell, and Deluca (1992), 52% of women reported having been sexually victimized at some time, which included attempted verbal or physical coercion to participate in sexual intercourse. In a study of high school students, it was found that 50% of females and 36.8% of males reported having suffered one form of sexual harassment (sexual comments), although some of the incidents reported by males were questionable as to whether they constituted sexual harassment as they referred to interactions between males that may not be considered sexual in nature, e.g., wedgies (Roscoe et al., 1994). In an occupational setting, figures were similar to those reported for university settings with 42% of females and 15% of males reporting that
Sexual Harassment
775
they had been sexually harassed as defined by the American Equal Employment Opportunity Commission. Behaviors included in this definition were unwelcome sexual advances and requests for sexual favors, and other physical and verbal conduct of a sexual nature (Charney & Russell, 1994). Thus, while prevalence rates differ from study to study, it is nonetheless safe to assume that the incidence of sexual harassment, at least in educational and employment contexts, is quite high. However, Charney and Russell (1994) reported that only 1-7% of people who experience sexual harassment file formal complaints, which indicates that official complaints of sexual harassment are not necessarily reflective of the frequency of this type of behavior. So why do people engage in sexual harassment? It may be argued that males, who are usually the perpetrators of this type of behavior (Charney & Russell, 1994), are carrying out behavior that they consider to be appropriate for their gender role. Thus, if males are rewarded for "sexual conquests" through admiration from their peers, it is not surprising that they may pursue such sexual activities more aggressively in order to obtain such rewards. As stated by Terpstra and Baker (1986, p. 19), "Sexual harassment is simply the exhibition of conditioned behaviors that are in accord with societal sex-role stereotypes." Pryor, Lavite, and Stoller (1993) have also noted that a link may exist between social dominance and male sexuality in the minds of men who are likely to sexually harass. Gagnon and Simon's (1973) script theory may be used to explain gender differences in sexually harassing behavior. According to Gagnon (1973), sexual scripts are learned in a similar manner as social scripts. Gagnon (1973) suggested that the script employed assists the actor to recognize what is potentially a sexual situation. Thus, some males may recognize situations as sexual when some women do not. Upon recognizing the potentially sexual nature of the situation, these males may enact what they consider to be appropriate and normal sexual scripts such as being assertive and making the first move. However, because the female does not perceive the situation to be sexual, the behavior exhibited by the male is often unwelcome and thus, may be considered sexual harassment. Significant gender differences exist in males' and females' perceptions of sexual harassment (Jones & Remland, 1992; Popovich, Gehlauf, Jolton, Somers, & Godinho, 1992; Powell, 1986; Tata, 1993). Tata (1993) reported that females were more likely to label sexual remarks and offensive jokes that occurred in the workplace as sexual harassment than were males. However, such gender differences disappeared when males' and females' perceptions of sexual bribery, sexual coercion, and sexual assault were compared, suggesting that more serious forms of sexual harassment are easier to recognize and define as such.
776
Foulis and McCabe
Popovich et al. (1992) also found that gender differences existed in perceptions of sexual harassment, regardless of the form that the sexual harassment took or the consequences of such behavior. Female raters tended to perceive the incident described as more likely to be sexual harassment, more likely to have an effect on the recipient of the harassing behavior, and more negative than did male raters. It was also found that males generally perceived the incident of harassment to be based on attraction, as opposed to females who perceived it to be more power based. Powell (1986) also found that females defined more incidents of behavior as sexual harassment, and Reilly, Lott, Caldwell, and Deluca (1992) found that college males were more tolerant of sexually harassing behavior than their female counterparts. This research suggests that females are more likely than males to perceive incidents as sexual harassment. Mazer and Percival (1989) also found that females had more negative attitudes to sexual harassment. They found that " . . . p e r c e p t u a l and attitudinal variables are negatively related to each o t h e r . . . t h o s e who define more incidents as harassment have less tolerance of it . . ." (Mazer & Percival, 989, p. 143). However, the differences found in these studies may be due to factors other than gender--for example, gender role or gender role stereotyping. Without evaluating these variables in the same study, the unique influence of gender is difficult to determine. The effect of gender role on attitudes to sexual harassment has received little attention. Bursik (1992) demonstrated that neither gender role nor gender were significantly related to perceptions of sexual harassment. Powell (1986) reported that those participants who scored high on femininity perceived more sexual harassment in some of the behaviors described, and those men who scored high on masculinity perceived less sexual harassment than men low on masculinity. On the other hand, women high on masculinity perceived the same behavior more as sexual harassment than women low on masculinity. Since there are limited studies on the impact of gender role on the perception of sexual harassment, the influence of this factor, and how it interacts with other variables in influencing perceptions of sexual harassment, requires further exploration. Gender role stereotypes, referring to stereotypic attitudes about what is appropriate and acceptable behavior for each gender, have received considerable attention as a variable that may impact on attitudes to sexual harassment. It must be noted that this variable differs from gender role in that gender role involves self-perception whereas gender role stereotypes refer to attitudes to stereotypic gender role behavior or what is considered to be male-appropriate and female-appropriate behavior. In Mazer and Percival's (1989) study, it was found that gender role stereotypes were significantly related to attitudes to sexual harassment with respondents who
Sexual Harassment
777
endorsed sexist attitudes being more accepting and tolerant of sexual harassment. It was also found that respondents with less sexist attitudes defined more incidents of behavior as sexual harassment. Similar results have been reported in relation to attitudes to sexual harassment using the Adversarial Sexual Beliefs Scale (Burt, 1980) with those participants who endorse adversarial sexual beliefs having more accepting attitudes to sexual harassment (Murrell & Dietz-Uhler, 1993; Reilly et al., 1992). Bartling and Eisenmann (1993) reported similar findings, and reported that those participants who had stronger feminist attitudes were less tolerant of sexual harassment. Similarly, Zaitchik and Mosher (1993) found that macho personality was the best predictor of sexual coercion/assault for men. Few studies have investigated the effect of age on attitudes to sexual harassment. ReiUy, Lott, and Gallogly (1986) found that younger university students were more accepting of sexual harassment than older university students. Lott et al., (1992) also reported that younger participants were more tolerant of sexual harassment than older participants. However, because age differences were generally represented by participants holding different positions within the university sample, that is, younger participants were students and older participants were staff and faculty, it is difficult to determine whether differences were attributed to age alone or the occupations of the participants. While a great deal is known about the experiences and attitudes of university students with respect to sexual harassment, little is known about younger adolescents' attitudes. Roscoe et al., (1994) found that among younger adolescents no gender differences existed in the degree of disapproval that participants had for sexually harassing behaviors. However, other studies have obtained different results. Feltey, Ainslie, and Geib (1991) found younger high school students were more likely to support sexually coercive behavior than older students. Although sexual coercion is different from sexual harassment, these behaviors also share some characteristics in that they are sexual, can be aggressive in nature, and are not welcomed by the targets of this behavior. Certainly, the attitudes of young adolescents toward sexual harassment needs further study. It is particularly important to determine how their responses are different to older respondnets on the same measures of sexual harassment. The variable of experience of sexual harassment has also received little attention. It may be expected, for example, that those participants who have experienced sexual harassment may be more disapproving of such behavior, and that this may be reflected in their attitudes and perceptions. However, Mazer and Percival (1989) did not find this to be the case. Instead, "Most of the perceptual and attitudinal variables investigated in this study were
778
Foulis and McCabe
found to be unrelated to level of sexual harassment experience" (p. 144). Similar findings were reported by Reilly et al. (1992). But are the findings on sexual harassment due to the nature of the workplace, and not the individuals concerned? Baker, Terpstra, and Cutler (1990) compared perceptions of sexual harassment between workers and university students, and found that significant differences existed between these two groups with a higher percentage of working women than student women perceiving the scenario described as sexual harassment. However, these groups were examined in two separate studies with three years elapsing between the studies. As a result, it is difficult to determine whether differences were due to occupation or to other variables that may have resulted from the time lapse. For example, sexual harassment laws may have changed during this period, or sexual harassment may have received more media attention during this time, and thus may have effected attitudes to sexual harassment. The current study was designed to explore the role of gender, gender roles, gender role stereotypes, age, and occupation on attitudes, perceptions, and experience of sexual harassment. Respondents from a range of situations (school, university, employment) were chosen to partcipate in the study. Although past research has indicated that a number of the above variables impact on sexual harassment in a partciular manner, they have not all been included in a single study, and so their unique contribution to the attitudes, perceptions, and experience of sexual harassment is unknown. From previous research findings it was expected that the following groups would report fewer perceptions of behavior as sexual harassment, more positive attitudes to sexual harassment, and fewer experiences of sexual harassment: males as opposed to females, respondents adopting a high masculine gender role as opposed to a high feminine gender role, respondents showing strong endorsement of sexist attitudes as opposed to egalitarian or feminist attitudes, younger as opposed to older respondents, students as opposed to workers.
METHOD
Subjects Participants consisted of three groups: 73 university students (27 males, 46 females) with a mean age of 20.54 years; 48 Year 10 secondary school students (13 males, 35 females) with a mean age of 15.56 years; and 75 office workers from all staff levels (33 males, 42 females) with a mean age of 34.1 years. The majority of participants were of Anglo-Saxon descent.
Sexual Harassment
779
Materials
1. The Sexual Harassment Experience Questionnaire (SHEQ) was designed to assess participants' experiences of sexual harassment over the past year and to ascertain the gender, age, and occupation of participants. Behaviors obtained from a leaflet titled "The Sex Discrimination Act 1984: Sexual Harassment--Knowing Your Rights" (Australian Human Rights and Equal Opportunity Commission, 1984) were listed, and participants were asked to indicate how many times such behaviors had been directed at them against their will in the past year. These behaviors were touching; patting or fondling; staring or leering; sexually suggestive comments and jokes; persistent invitations or requests for sex; showing lewd pictures, posters, or cartoons; and unnecessary familiarity, such as deliberately brushing up against someone. The range of possible scores on this questionnaire is 6-24 with lower scores indicating less experience of sexual harassment. The four response categories for levels of experience of sexual harassment and scoring key are "never" (1), "1-2 times" (2), "3-4 times" (3) and "5+ times" (4). 2. The Sexual Harassment Attitude Scale (SHAS) is a 19-item scale constructed by Mazer and Percival (1989) to assess attitudes to sexual harassment and is an extension of the Tolerance for Sexual Harassment Inventory (TSHI; Lott et al., 1982; Reilly et al., 1986). The five response categories and scoring key are "strongly agree" (5), "agree" (4), "undecided" (3), "disagree" (2), and "strongly disagree" (1) with items 4 and 7 reverse coded. The range of possible scores for this scale are 19-95 with higher scores indicating more tolerant attitudes to sexual harassment. The SHAS Coefficient alpha of .84 indicates high internal consistency and the additional items of the SHAS are strongly related to the TSHI (r = .61, p < .001; Mazer & Percival, 1989). 3. The Australian Sex Role Scale devised by Antill, Cunningham, Russell, and Thompson (1981) has 50 items measuring the masculinity and femininity of characteristics attributed to people, and asks participants to rate to what extent such characteristics could be used to describe themselves. The seven response categories range from "never or almost never" (1), "usually not true" (2), "sometimes but infrequently true" (3), "occasionally true" (4), "often true" (5), "usually true" (6), and "always or almost always true" (7). The range of possible scores for both masculinity and femininity is 20-140. Higher scores indicate high masculinity or femininity. The coefficient alphas for the four main scales were .67 (M+), .78 (M-), .75 (F+), and .67 (F-) (Marsh, 1987). 4. The Macho Scale devised by Villemez and Touhey (1977) is a 28item scale designed to assess stereotypical beliefs about male and female
Foulis and McCabe
780
appropriate behavior. High scores on this scale indicate more sexist beliefs about male and female behavior. The response categories and scoring key were "strongly agree" (5), "agree" (4), "undecided" (3), "disagree" (2), and "strongly disagree" (1). The range of possible scores on this scale is 28-140 with higher scores indicating more sexist attitudes. The Macho scale has test-retest correlations ranging from .89 to .94 (Villemez & Touhey, 1977). 5. The final questionnaire was titled the Sexual Harassment Definitions Questionnaire (SHDQ). Vignettes were constructed based on the behaviors listed in the SHEQ, which were drawn from the definition of sexual harassment adopted by the Australian Human Rights and Equal Opportunity Commission (1984). For each incident of possible sexual harassment, four vignettes were created--two same gender incidents and two opposite gender incidents with the gender of the perpetrator and recipient being alternated (see Appendix A). This made it possible to determine whether the gender of the perpetrator or recipient affected definitions of sexual harassment. The response categories and scoring key for defining sexual harassment is "Yes" (1) and "No" (0) with possible scores ranging from 0 to 16. A dichotomous variable was used to simplify the administration of the scale and to force respondents to make a decision regarding whether or not they perceived the incident to involve sexual harassment. Higher scores indicate that participants are defining more incidents as sexual harassment.
Procedure Group 1. Once ethical approval had been obtained, advertisements explaining the purpose of the study and what was required of participants were placed on notice boards at two campuses of the University. Students were assured that they would remain anonymous and their individual results would be confidential. Volunteers took no longer than 40 minutes to complete the questionnaires and they completed the questionnaires under classroom conditions. Most respondents were born in Australia, with 25% of respondents' parents born overseas, mostly from European countries. Group 2. After consent was obtained from the Directorate of School Education, the randomly selected schools within the metropolitan area of a large city were contacted via letter requesting assistance with the study. Two of these schools agreed to take part in the study. Parents of the students were sent the plain language statement and consent form, which stated that participants would remain anonymous and that individual results would be kept confidential. Students participated only after parental consent had been obtained. Once this procedure had taken place, students completed the questionnaires during class time. Seventy-five percent of re-
Sexual Harassment
781
spondents were born in Australia, with 45% of respondents' parents born overseas, mostly from European countries. Group 3. Once consent from the employer of a company with more than 100 employees had been obtained, questionnaires were distributed to employees with the plain language statement and consent form. This employer was selected because it was a large company with a large number of male and female office staff. This was the only company selected to take part in the study. There was an Equal Opportunities Officer employed by the company, but there had been no courses run by the company on sexual harassment. The workers were informed that they would remain anonymous and that their individual results would be kept confidential. Employees completed the questionnaires during work time and returned them to the liaison person at their place of employment. The questionnaires were kept separate from the consent forms and were submitted in a plain envelope. Eighty-five percent of respondents were born in Australia, and 60% of respondents' parents were born in Australia.
RESULTS There were 196 participants (73 males, 123 females) from whom data could be used. The data were screened for outliers and multicolinearity. No variables were shown to exceed a Z score of 3, and an examination of the correlation matrix showed an absence of singularity and multicolinearity. Attitudes to Sexual Harassment
In order to determine the gender differences between each of the three groups of respondents on attitudes to sexual harassment, an analysis of variance was conducted, with the dependent variable being respondents' scores on the SHAS and the independent variables being gender and group. There was a significant main effect for group [F(2,193) = 19.96, p < .001], gender [F(2,193) = 24.13, p < .001], and for the Group • Gender interaction [F(2,193) = 4.71, p < .01]. Summary statistics for attitudes to sexual harassment are provided in Table I. Results of t tests indicated that, in general, males and females differed significantly in their attitudes to sexual harassment, with males being more accepting of sexual harassment than females (t = 3.30, p < .001). This difference was mainly produced by the difference in males' and females' attitudes to sexual harassment in Group 2 (t = 4.48, p < .001). Gender differences on the same variable approached significance for Group 1 (t = 1.91, p = .060), whereas there was no significant gender
782 Table
Foulis and McCabe
I. Male and Female Mean Scores on the Sexual Harassment Attitude Scale According to Group Males
All participants (n = 196) Group 1 (n = 73) Group 2 (n=48) Group 3 (n = 75)
52.91 n = 73 49.90 n = 27 66.32 n = 13 50.09 n = 33
Females 47.85 n = 123 45.72 n = 46 51.93 n = 35 46.78 n = 42
t values
Group Means
3.30a 1.91
47.26
4.48a
55.83
1.46
48.24
F = 11.72a
ap < .001.
difference in sexual harassment attitudes for Group 3. Respondents from Group 2 demonstrated more tolerant attitudes to sexual harassment than those from Group 3 (t = 7.15, p < .01) and Group 1 (t = 7.35, p < .01), but there were no other significant differences between the groups. A correlation revealed no significant relationship between age and attitudes to sexual harassment, although it did approach significance (r = -.14, p = .055), suggesting that the younger participants were more accepting of sexual harassment. This is supported by the fact that the high school students (the youngest group) scored significantly higher than the other groups on the SHAS. Multiple regression analyses were also conducted on each group with attitudes to sexual harassment entered as the dependent variable, and gender role scores, sexual stereotype scores, experience of sexual harassment, perceptions of sexual harassment, and occupation entered as prediction variables. The results are shown in Table II. Overall, the equation was highly significant (F = 22.93, p < .001). The variables that contributed most to attitudes to sexual harassment were scores on the Macho scale, perceptions, experience of sexual harassment, and masculinity. These resuits were separately analysed for each group. All three analyses were highly significant (F = 16.70,p < .001, for Group 1; F = 12.90, p < .001, for Group 2; F = 7.67, p < .001, for Group 3). For both student groups, the predictor variables accounted for a high amount of the variance (Adjusted R 2 Group 1 = .52; Group 2 = .56), with the scores on the Macho scale contributing the most variance to attitudes to sexual harassment in both cases (F = 51.96, p < .001, for Group 1; F = 49.84, p < .001, for Group 2). However, in Group 1, perceptions and scores on the masculinity subscale also contributed uniquely to the variance, while this did not occur for Group 2. For Group 3, the predictor variables accounted for only 29% of the variance in scores on the SHAS with perceptions contributing the
Sexual Harassment
783
Table II. Multiple Regressionwith Scoreson Macho Scale, Perceptions,Experience, Gender Role, Occupation, and Gender Entered as Predictors of SHAS Scores for Each Group Variables in Equation Macho scale (Gender role stereotyping) Perceptions Experience Femininity Masculinity Occupation (Group) Gender Multiple R R2 Adjusted R 2
Standard Error ~pp< .05. < .01. Cp < .001.
Beta .51 -.23 .17 -.01 .12 .01 -.09
sr2
F
.18 .04 .01 .00 .01 .00 .00
75.14c 16.21Y' 6.62b .04 5.10a .06 2.12
.68 .47 .44 7.84
F = 22.93 c
most unique variance (F = 19.92, p < .001), with scores on the Macho scale also being significant (F = 7.58, p < .01). Scores on the femininity subscale also significantly contributed to the variance (F = 4.26, p < .05). Thus, factors contributing to attitudes to sexual harassment varied according to each group.
Perceptions of Sexual Harassment Summary statistics are provided in Table III. Results of t tests revealed no significant gender differences in perceptions of sexual harassment for the sample. This also applied for Groups 1 and 3. Group 2, however, approached significance with males defining more incidents as sexual harassment than females (t = -1.07, p = .055). A correlation revealed no significant relationship between age and perceptions of sexual harassment. There were no significant group differences. Multiple regression analyses were performed on each group's perceptions of sexual harassment with scores on the Macho scale, gender role scale, attitudes to sexual harassment, and experience of sexual harassment entered as predictor variable. The results are shown in Table III. The results demonstrated that this set of varialbes significantly predicted perceptions of sexual harassment (F = 5.25, p < .001), with attitudes to sexual harassment providing unique prediction (F = 16.20, p < .001). The results suggest that different factors are impacting on perceptions according to group. While attitudes contributed uniquely to the total variance for both Group 1 (F = 16.20, p < .001) and Group 3 (F = 19.92,
Foulis and McCabe
784
Table III. Multiple Regression with Scores on the Macho Scale, Attitudes to Sexual Harassment, Experience of Sexual Harassment, Gender Role, Occupation, and Gender Entered as Predictors of Perceptions Variables in Equation Macho scale Attitudes Experience Femininity Masculinity Occupation Gender Multiple R R2 Adjusted R 2 Standard Error
Beta -.04 -.35 -.07 .08 .03 -.05 .01 .41 .17 .13 4.56
sr2
F
.00 .08 .00 .00 .00 .00 .00
.22 t6.20a .76 1.34 .21 .21 .02
F = 5.25 '~
ap < .001.
p < .001), this was not the case for Group 2, which had femininity as the only unique predictor (F = 8.44, p < .01).
Experience of Sexual Harassment In order to determine the gender differences between each of the three groups of respondents on experience of sexual harassment, an analysis of variance was conducted, with the dependent variable being respondents' scores on the SHEQ and the independent variables being gender and group. There was a significant main effect for group [F(2,193) = 5.80, p < .001], gender [F(1,193) = 5.32, p < .001], but not for the Group x Gender interaction. Results of t tests (see Table IV) indicate that significant gender differences existed in participants' experience of sexual harassment for the whole sample with females experiencing more sexual harassment in the past year than their male counterparts (t = -3.10, p < .01). Gender differences were not found for Groups 1 and 2, but did exist for Group 3 (t = -3.64, p < .001). On all types of sexual harassment, females cited having experienced more incidents than males. It was also found that friends, peers, and strangers were all perpetrators of sexual harassment and that the sexual harassment occurred most often in a social setting. Students experienced more sexual harassment than workers, with there being significant differences between respondents in Group 1 and Group 3 (t = 7.21, p < .01) and Group 2 and Group 3 (t = 6.94, p < .01). A correlation between age and experience of sexual harassment revealed a strong negative relationship (r = -.55, p < .001), indicating that experience of sexual harassment decreased as age increased.
Sexual Harassment
785
Table IV. Male and Female Mean Scores for Experience of Sexual Harassment According to Group With Age Included as a Covariate Males All participants (n = 196) Group 1 (n = 73) Group 2 (n=48) Group 3 (n = 75)
9.44 n = 73 11.04 n = 27 12.69 n = 13 6.85 n = 33
Females 11.40 n = 123 12.85 n = 46 12.69 n = 35 8.74 n = 42
t values
Group Means
-3.10 a -1.75
12.18
.00
12.69
-3.64 b
7.91
F = 18.36 b p<
.01. < .001.
A linear multiple regression was performed on the data with experience of sexual harassment entered as the dependent variable and attitudes to sexual harassment, perceptions of sexual harassment, scores on the Macho scale, gender role scores, occupation, and gender entered as the predictor variables. The results are displayed in Table V below. This analysis was highly significant (F = 15.33, p < .001), with variables accounting for 34% of the combined variance. Occupation contributed the most unique variance (F = -7.02, p < .001) with gender (F = 4.46, p < .001), attitudes (F = 2.57, p < .01), and masculinity (F = 3.76, p < .001) also contributing to the variance.
DISCUSSION High school males had more tolerant attitudes toward sexual harassment than females. However, there were no differences between males and females at university and in the workplace in their attitudes. The findings in relation to the high school students are consistent with the results from other studies (Mazer & Percival, 1989) and offer support for these theoretical frameworks regarding gender differences. As suggested by social learning and script theory, people become gender typed at an early age as a result of reinforcement and modeling behavior on same-gender role models (Gagnon & Simon, 1974; Mischel, 1966). Gender differences occur as a result of this conditioning and these differences influence attitudes to sexual harassment. Males perceive more situations as being sexual or potentially sexual, and so view sexually harassing behavior as normal or appropriate. For example, they may see sexual harassment as normal flirtation between men and women.
786
Foulis and McCabe Table V. Multiple Regression with Scores on the Macho Scale, Attitudes to Sexual Harassment, Perceptions, Gender Role, Occupation, and Gender Entered as Predictors of Experience of Sexual Harassment
Variables in Equation Macho scale Attitudes Perceptions Femininity Masculinity Occupation Gender Multiple R R2 Adjusted R 2 Standard Error
Beta -.14 .18 -.06 .09 .21 -.42 .29 .61 .37 .34 .14
sr2
F
.00 .02 .00 .00 .04 .17 .05
-.91 2.57a -.87 .12 3.76b -7.02 b 4.46b
F = 15.33b
< .001. ~ P < .001.
However, the fact that high school students were the only group to exhibit gender differences in their sexual harassment attitudes would not appear to be consistent with these theories. One would expect gender differences to be stronger in the workers' attitudes because this group is older than the other groups, and gender roles would have been far more rigid when this group went through the gender-typing process during childhood. Because gender roles are now far more flexible and have certainly been so for the last twenty years, it could be expected that the youngest group would not exhibit gender differences in attitudes as was found by Roscoe et al. (1994). Alternatively, it could be argued that as people's experience of the world increases, they incorporate what they have learnt into their schemas or scripts as suggested by Nelson and Keith (1990). This explanation would support the decline of gender differences with an increase in age. The workers were older than the other two groups and the results indicated that older males and females shared more similar views on gender-appropriate behavior than their younger counterparts. High school students were found to have more tolerant attitudes to sexual harassment than either of the other groups, suggesting that acceptance of sexually harassing behavior decreases as age increases or, alternatively, that increasing exposure to the work force is associated with less acceptance of such behavior. Whether this finding is due to age or occupation, it indicates that experience of the world has an impact on attitudes to sexual harassment. However, occupation was not found to be a predictor of attitudes to sexual harassment, nor was it found to correlate with attitudes. This is inconsistent with the above findings. A possible interpretation of this finding is that the group differences are age related rather than
Sexual Harassment
787
related to occupation. A correlation between age and attitudes to sexual harassment approached significance, indicating that this is likely to be the case. Past research also suggests that age influences attitudes to sexual harassment, with younger participants being more tolerant and accepting of this behavior (Reilly et al., 1986). Gender role may be crucial to the explanation of the high school students' more tolerant attitudes to sexual harassment, as it was found the high school students scored significantly higher on masculinity than the other two groups. Masculinity was also found to predict attitudes to sexual harassment for the whole sample and for the university students, and to a lesser extent, high school students. For workers, however, it was femininity that acted as a predictor. Previous research has demonstrated that gender role has some effect on perceptions, and it seems that gender role may influence attitudes in a similar way. However, in past studies on perceptions, researchers have failed to offer any explanation for such findings (Powell, 1986). In the current study, gender role appears to be impacting on attitudes in very different ways for each group and may be a reflection of the age differences and life experiences of the different groups. The strongest predictor of attitudes to sexual harassment was gender role stereotypes, as revealed through Macho scale scores. This indicates that, as expected, sexist attitudes and attitudes to sexual harassment were strongly linked. The more sexist a person was, the more accepting they were of sexual harassment. Both sets of attitudes may result from the gender-typing process. Tolerance of sexual harassment may be interpreted as a form of sexism in itself with those participants possessing more sexist attitudes being more inclined to perceive sexual aggression as male-appropriate behavior, and to perceive resistance to this sexual aggression as female-appropriate behavior. This relationship between attitudes and gender role stereotypes is also explained by script theory (Gagnon, 1973), which suggests that acceptance of sexual harassment may reflect the adherence of participants to scripts associated with gender role stereotypes. Interestingly, university students had significantly lower scores on the Macho scale, indicating that they had the least sexist attitudes. From an age perspective, this is unusual because this group was the middle group in terms of age. However, such a finding could be occupation related, with university students being exposed to greater equality between the genders at university. Participants' perceptions of sexual harassment were found to significantly predict attitudes to sexual harassment. The correlation between these two variables revealed that those participants with more tolerant attitudes toward sexual harassment perceived less incidents to be sexual harassment. This finding is consistent with past research in which perceptions and attitudes were found to be negatively correlated (Mazer & Percival, 1989).
788
Foulis and McCabe
Perceptions played a more important role in attitudes to sexual harassment for the workers than gender role stereotypes, with perceptions accounting for a major proportion of the variance. This may be due to the fact that workers have had a greater experience of sexual harassment, and so their perceptions of what constitutes sexual harassment are more likely to influence their attitudes toward this type of behavior. Attitudes to sexual harassment also were found to be a predictor of perceptions for university students and workers, with this variable playing a major role for the latter group. Both the workers' and university students' attitudes predicted their perceptions while this was not the case for the high school students. Older respondents may have had more opportunities to test their attitudes against their perceptions and thus may have adjusted their attitudes when conflicts arose. Younger respondents, however, have had limited experience and so any conflict between attitudes and perceptions has yet to occur. There were no gender differences found in perceptions for the whole sample or for each group. This is an unexpected result as it was predicted that gender differences would exist. However, the present study was conducted with participants from Australia, whereas other studies were this research was conducted in America, and so inconsistencies may be the product of cultural differences. It is also surprising because attitudes were affected by gender and one would expect that attitudes would effect the way in which the various situations were perceived. However, this finding is consistent with the fact that there was only a minor relationship between attitudes and perceptions. These results suggest that different factors are impacting on attitudes and perceptions. It appears that males and females both recognize incidents as sexual harassment but differ in how they rate the incident in terms of acceptability. One possible explanation for this finding may be that, due to increased media attention on sexual harassment during the period in which the study was conducted, participants may have been more aware of what behaviors were defined as sexual harassment by law, and thus defined incidents in accordance with those behaviors. Thus, males and females perceived incidents of sexual harassment in a similar manner, both defining such incidents according to their understanding of legal definitions or even according to the way in which they thought society would define such incidents. Femininity was found to play an important role in perceptions for the high school students and was, in fact, the only predictor of perceptions for this group. It did not, however, have an effect on the other groups or on the sample as a whole. The large role of femininity on perceptions for this group may result from the limited experience these participants have of the world. The other groups, due to their greater experience, had many
Sexual Harassment
789
more factors influencing their perceptions, and so the effect of gender role was decreased. Responses on the Macho scale did not significantly predict perceptions, although they played a strong role in attitudes. This is consistent with the findings of Mazer and Percival (1989), who found a significant correlation between gender role stereotypes and perceptions, but no predictive relationship was found. This again offers support for the probability that different factors from attitudes are leading to perceptions of sexual harassment. Sexist attitudes and attitudes to sexual harassment may be so strongly related because they share common attitudinal factors and both required value judgments on the acceptability of behaviors. Alternatively, perceptions involve respondents defining behavior, regardless of how they feel about the appropriateness of this behavior. Consistent with past research (Lott et al., 1992), females experienced more sexual harassment than males. However, such differences did not exist for the students, and may indicate that younger females, in accordance with changing gender roles, are becoming more sexually aggressive. However, students experienced higher levels of sexual harassment than the workers. This may be due to the fact that the students are not likely to be married and thus are more likely to be looking for, and actively seeking a sexual relationship, placing themselves in situations which may lead to sexual harassment. Results indicated that prevalence of sexual harassment for males and females was high, with over a third of males having experienced sexual harassment and approximately half of females also having experienced sexual harassment. These findings are consistent with past research conducted by Roscoe et al. (1994) on adolescent experiences. However, in studies conducted on university students (Lott et al., 1982) and workers (Charney & Russell, 1994), prevalence of sexual harassment for males was not found to be this high. This difference in results is most likely a reflection of the different definitions of sexual harassment used between studies. Both occupation and age were found to impact on experience of sexual harassment, with occupation being a significant predictor of experience and age correlating quite highly with this variable. Workers were least likely to experience sexual harassment, and sexual harassment was found to decrease with an increase in age. Thus, it appears that sexual harassment may be a more persistent problem for younger people. One explanation for this finding is that younger people are more likely to be unmarried and this may make sexual harassment more acceptable to the actual harassers. It is also possible that younger respondents were more susceptible to sexual harassment. Attitudes were found to bc a predictor of experience of sexual harassment with respondents who had experienced sexual harassment being more tolerant of this type of behavior. While it could be expected that
790
Foulis and McCabe
respondents would be less tolerant if they had experienced more sexual harassment, it may be that respondents have become more accustomed to this behavior and see it as inevitable. Alternatively, when respondents were asked about their experiences, they were not required to label them as sexual harassment and thus they may have perceived them to be normal interactions. Again, it must be emphasized that even when behavior is labeled as sexual harassment, this does not automatically mean that the behaviors are considered to be unacceptable. Masculinity was also found to predict experience of sexual harassment and may reflect the occupational differences found in regard to this variable. Both student groups were higher in masculinity than the workers and higher in experience of sexual harassment. Thus, this effect may be more a consequence of age rather than any relationship between masculinity and experience of sexual harassment and this seems to be the case.
SUMMARY The overall findings suggest that while attitudes and perceptions of sexual harassment are related, they also differ, in that attitudes require value judgments to be made of behavior. This explains the high relationship between attitudes to sexual harassment and attitudes regarding gender role stereotypes. Sexist attitudes are associated with acceptance of sexual harassment. The impact of age and occupation on attitudes suggest that high school students adhere to gender role stereotypes and have a high tolerance of sexual harassment. However, this group's lack of experience in the world and, consequently, their limited exposure to attitudes that may challenge their current scripts and schemas, may be responsible for these findings. In general, the university students held less sexist attitudes about male and female appropriate behavior and, accordingly, had a lower tolerance for sexual harassment. This was attributed partly to occupation, as it appears that the university environment may challenge preconceived ideas about appropriate gender role behavior and the students may have incorporated these new experiences into their schemas. The workers, while having had more experience of the world than the other groups, still adhere to the gender role stereotypes with which they were raised and socialised. Overall, males were more tolerant of sexual harassment which would suggest that males are socialized to be the sexual aggressors and to perceive situations as being more sexual than females. Different factors impacted on attitudes and perceptions of sexual harassment. There were no gender differences in perceptions, thus indicating
Sexual Harassment
791
that males and females generally perceived the same incidents as sexual harassment. Differences between this study and previous studies that have found gender differences in perceptions may be attributed to cultural differences and may also be a consequence of media exposure. In the past year, in Australia, sexual harassment has received increased media attention and awareness of sexual harassment issues may have been raised, resulting in more consistent perceptions between males and females. Females tended to experience more sexual harassment than males. Males are socialized to be the sexual aggressors while sexual aggression in females is not considered desirable. Thus, males tend to sexually harass females more than females do males. These findings suggest that inconsistencies still exist between males' and females' attitudes to sexual harassment about what is acceptable behavior. As long as males continue to have more tolerant attitudes of sexual harassment, perhaps resulting from confusion about what constitutes appropriate gender role behavior for various situations, the problem of sexual harassment is likely to continue. Education is necessary to decrease sexual harassment, either through campaigns to raise awareness about what behaviors constitute sexual harassment, or through education directed at decreasing stereotypic beliefs. If gender role stereotypes are reduced, a reduction in the adherence to scripts associated with these stereotypes may result, thus decreasing sexually harassing behavior as a consequence. The results of the current study also suggest that this type of education may be of most use at the high school level. The incidence of sexual harassment may also decrease naturally with gender role stereotypes continually being challenged. Changes in the gender role models that parents and the media exhibit may be reflected in decreasing stereotypic attitudes in adolescents, and consequently less acceptance of sexual harassment. This study needs to be replicated with a larger, more representative sample. Ethnicity also needs to be addressed and participants of the same age sampled from the various sections of the community, so that the effects of age and occupation can be investigated and interpreted separately.
APPENDIX A Sexual Harrassment Definitions Questionnaire
Instructions
Please read these descriptions of behaviors and tick the appropriate boxes. (You may tick more than one box if you feel that this is necessary.)
792
Foulis and McCabe
1. Jane is walking along the street when two men yell sexually suggestive comments to her from a passing car. If you were Jane, would you be: flattered
I
I
offended
I~1
upset
I
angry
I
annoyed not bothered
I I
I
I
I
I
Would you define this incident as sexual harassment? Yes 2. When at work, a female colleague deliberately brushes up against Mike when walking past, even though there is no need for this as there is plenty of room. If you were Mike, would you be: flattered
I
I
offended
I
I
upset
I
angry
I
I
annoyed not bothered
I
I I
I
I
Would you define this incident as sexual harassment? Yes[
[
No[
[
3. Jane is at a party when a women whom she has never met squeezes her on the bottom. If you were Jane, would you be: flattered
I
offended
I I
upset
annoyed not bothered
angry
I I----1
I
I I
I
Would you define this incident as sexual harassment?
I
I
Nol
l
V---]
793
Sexual Harassment
4. While at work, a couple of female colleagues make a point of showing Mike a sexually explicit cartoon in which a male is depicted in a derogatory way. If you were Mike, would you be: flattered
I
I
offended
I
I
upset
I
angry
I
I
annoyed not bothered
I I
I
iml
Would you define this incident as sexual harassment?
~es
I
I
No I
I
5.When at work, a male colleague deliberately brushes up against Mike when walking past, even though there is no need for this as there is plenty of room. If you were Mike, would you be: flattered
I
I
offended
I
I
upset
angry
F--q
I
annoyed not bothered
I I
I
I
I
Would you define this incident as sexual harassment?
~os
I
I
~o
I~1
6. A male acquaintance of Jane's has asked her out a number of times and each time she has refused to go out with him. However, he continues to persist in his requests for a date. If you were Jane, would you be: flattered
I
offended
I I
upset
I I~1
angry
I
annoyed not bothered
I I~1
Would you define this incident as sexual harassment?
I
I
794
Foulis and McCabe
7. Mike is walking along the street when two women yell sexually suggestive comments to him from a passing car. If you were Mike, would you be: flattered
I
I
offended
I
I
upset
I
angry
I
I
annoyed not bothered
I I
I
I
I
Would you define this incident as sexual harassment?
~os
I
I
~o[~1
8. While at work, a couple of male colleagues make a point of showing Jane a sexually explicit cartoon in which a female is depicted in a derogatory way. If you were Jane, would you be: flattered
I
offended
upset
I1~11
angry
I
I
annoyed not bothered
I I
I
I
I
Would you define this incident as sexual harassment?
Yes
I
I
No I
9. Jane is at a party when a man whom she has never met squeezes her on the bottom. If you were Jane, would you be: flattered
I
I
offended
I
I
upset
I
angry
I
I
annoyed not bothered
I I
I
I
Would you define this incident as sexual harassment?
Yes I
I
No I
I
I
Sexual Harassment
795
10. A female acquaintance of Mike's has asked him out a number of times and each time he has refused to go out with her. However, she continues to persist in his requests for a date. If you were Mike, would you be: flattered
I
offended
I I
upset
angry
I [~l
I
annoyed not bothered
I [~l
I
I
Would vou define this incident as sexual harassment?
Y~
I
I
~o
[~l
11. When at work, a female colleague deliberately brushes up against Jane when walking past even though there is no need for this as there is plentry of room. If you were Jane, would you be: flattered
I
I
offended
[~l
upset
I
angry
I
I
annoyed not bothered
I I
I
I
I
Would you define this incident as sexual harassment? Yes
[~l
No
[ ~
12. Mike is walking along the street when two men yell sexually suggestive comment to him from a passing car. If you were Mike, would you be: flattered
I
I
offended
I
I
upset
I
angry
I
I
annoyed not bothered
I I
I
I
I
Would you define this incident as sexual harassment?
~os
I
I
~o
[~
13.Mike is at a party when a woman whom he has never met squeezes him on the bottom.
Foulis and McCabe
796
If you were Mike, would you be: flattered
I
!
offended
!
I
angry
upset
I i
I
annoyed not bothered
t V--q
V---q
Would you define this incident as sexual harassment? Yes[
[
No[
t
14. Jane is walking along the street when two women yell sexually suggestive comments to her from a passing car. If you were Jane, would you be: flattered
I
I
offended
i
upset
angry
. .1. I ....I.... I
annoyed not bothered
! I
i
I----I
Would you define this incident as sexual harassment?
V--q
Nol
I
15. When at work, a male colleague deliberatley brushes up against Jane when walking past even though there is no need for this as there is plenty of room. If you were Jane, would you be: flattered
I
I
offended
I
I
upset
angry
I ....I... I
annoyed not bothered
I
I
I
I
Would you define this incident as sexual harassment?
16. Mike is at a party when a man whom he has never met squeezes him on the bottom. If you were Mike, would you be: flattered
I
i
offended
I
upset
I
angry
I
I
annoyed not bothered
I I
I
/----1
Sexual Harassment
797
W o u l d y o u d e f i n e this i n c i d e n t as sexual h a r a s s m e n t ? Yes]
]
No]
REFERENCES Antill, J. K., Cunningham, J. D., Russell, G., & Thompson, N. L. (1981). An Australian sex role scale. Australian Journal of Psychology, 33, 169-183. Australian Human Rights and Equal Opportunity Commission. (1984). The sex discrimination act 1984: Sexual harassment--knowing your rights. Australia: Australian Human Rights and Equal Opportunity Commission. Baker, D. D., Terpstra, D. E., & Cutler, B. D. (1990). Perceptions of sexual harassment: A re-examination of gender differences. The Journal of Psychology, 124, 409-416. Bartling, C. A., & Eisenman, R. (1993). Sexual harassment proclivities in men and women. Bulletin of the Psychonomic Society, 31, 189-192. Burt, M. R. (1980). Cultural myths and supports for rape. Journal of Personality and Social Psychology, 38, 217-230. Charney, D. A., & Russell, R. C. (1994). An overview of sexual harassment, American Journal of Psychiatry, 151, 10-17. Feltey, K. M., Ainslie, J. J., & Geib, A. (1991). Sexual coercion attitudes among high school students--the influence of gender and rape education. Youth and Society, 28, 229-350. Gagnon, J. H (1973). Scripts and the coordination of sexual conduct. Nebraska Symposium on Motivation, 21, 27-59. Gagnon, J. H., & Simon, W. (1973). Sexual conduct." The social sources of human sexuality. New York: Aldine Publishing Company. Jones, T. S., & Remland, M. S. (1992). Sources of variability in perceptions of and responses to sexual harassment. Sex Roles, 27, 121-142. Lott, B., Reilly, M., & Howard, D. R. (1982). Sexual assault and harassment: A campus community case study. Signs: Journal of Women in Culture and Society, 8, 296-319. Marsh, H. W. (1987). Masculinity, femininity and androgyny: Their relations with multiple dimensions of self-concept. Multivariate Behavioral Research, 22, 91-118. Mazer, D. B., & Percival, E. E (1989). Ideology or experience? The relationships among perceptions, attitudes, and experience of sexual harassment in university students. Sex Roles, 20, 135-147. Mischel, W. (1966). A social learning view of sex differences in behavior, in E. E. Maccoby (Ed.), The development of sex differences. Palo Alto, CA: Stanford University Press. Murrell, A. J., & Dietz-Uhler, B. L. (1993). Gender identity and adversarial sexual beliefs as predictors of attitudes toward sexual harassment. Psychology of Women Quarterly, 17, 169175. Nelson, C., & Keith, J. (1990). Comparisons of female nd male early adolescent sex role attitude and behavior development. Adolescence, 15, 183-204. Popovich, P M., Gehlauf, D. N., Jolton, J. A., Somers, J. M., & Godinho, R. M. (1992). Perceptions of sexual harassment as a function of sex of rater and incident form and consequence. Sex Roles, 27, 609-625. Powell, G. N. (1986). Effects of sex role identity and sex on definitions of sexual harassment. Sex Roles, 14, 9-19. Pryor, J. B., Lavite, C. M., & Stoller, L. M. (1993). A social psychological analysis of sexual harassment: The person situation interaction. Journal of Vocational Behavior, 42, 68-83. Reilly, M. E., Lott, B., Caldwell, D., & DeLuca, L. (1992). Tolerance for sexual harassment related to self-reported sexual victimization. Gender and Society, 6, 122-138. Reilly, M. E., Lott, B., & Gallogly, S. M. (1986). Sexual harassment of university students. Sex Roles, 15, 333-358. Roscoe, B., Strouse, J. S., & Goodwin, M. P (1994). Sexual harassment: Early adolescents' self-reports of experiences and acceptance. Adolescence, 29, 515-523.
798
Foulis and McCabe
Rubin, L. J., & Borgers, S. B. (1990). Sexual harassment in universities during the 1980s. Sex Roles, 23, 397-411. Tata, J. (1993). The structure and phenomenon of sexual harassment: impact of category of sexually harassing behavior, gender, and hierarchical level. Journal of Applied Social Psychology. 23, 199-211. Terpstra, D. E., & Baker, D. D. (1986). A framework for the study of sexual harassment. Basic and Applied Social Psychology, 7, 17-34. Villemez, W. J., & Touhey, J. C. (1977). A measure of individual differences in sex stereotyping and sex discrimination: The 'macho' scale. Psychological Reports, 41, 411-415. Zaitchik, M. C., & Mosher, D. L. (1993). Criminal justice implications of the macho personality constellation. Criminal Justice and Behavior, 20, 227-239.