17
Sex Roles, VoL 27, Nos. 11112, 1992 Contrapower Sexual Harassment: The Effects of Student Sex and Type of Behavior on Faculty Perceptions 1 Kathleen McKinney Illinois State The purpose of this study was to assess the effects of sex of offender and type of behavior on faculty perceptions of contrapower sexual harassment. Probability samples of predominately white male and female faculty at two universities in the Midwest were asked to read and make judgments about an incident that might constitute sexual harassment of a faculty member by an opposite-sex student. The effects of offender/student sex, type of harassing behavior, and subject sex were assessed. Some of the results included, as hypothesized, that when the student offender was male, subjects were more likely to: (1) view the incident as harassment, (2) believe the faculty member wo[tld be upset, and (3) see the student as responsible, compared to when the student offender was female. Overall, obscene phone calls and explicit verbal-physical harassment were viewed more negatively than written sexual comments and implicit verbal-physical harassment. Female faculty subjects also viewed the incidents as more negative or problematic on several measures than did male faculty. Numerous studies exist on sexual harassment in academia. Almost all of this research, however, deals with the sexual harassment of students by fac- ulty or with the harassment of faculty by other faculty (e.g., Adams, Kottke, & Padgitt, 1983; Benson & Thomson, 1982; Mazer & Percival, 1989; McKinney, 199Oa,b; McKinney & Maroules, 1991; McKinney, Olson, & Satterfield, 1988; IThe author would like to thank James Sikora for his assistance in collecting the data, and Susan Sprecher and Elizabeth Grauerholz for their comments on earlier versions of this paper. This research was funded by a University Research Grant from the Office of Research and Sponsored Programs at Illinois State University. 627 0 1992 Plenum Publishing Corporation

Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Embed Size (px)

Citation preview

Page 1: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Sex Roles, VoL 27, Nos. 11112, 1992

Contrapower Sexual Harassment:The Effects of Student Sex andType of Behavior on Faculty Perceptions1

Kathleen McKinneyIllinois State Uni~'ersity

The purpose of this study was to assess the effects of sex of offender and typeof behavior on faculty perceptions of contrapower sexual harassment.Probability samples of predominately white male and female faculty at twouniversities in the Midwest were asked to read and make judgments about anincident that might constitute sexual harassment of a faculty member by anopposite-sex student. The effects of offender/student sex, type of harassingbehavior, and subject sex were assessed. Some of the results included, ashypothesized, that when the student offender was male, subjects were morelikely to: (1) view the incident as harassment, (2) believe the faculty memberwo[tld be upset, and (3) see the student as responsible, compared to when thestudent offender was female. Overall, obscene phone calls and explicitverbal-physical harassment were viewed more negatively than written sexualcomments and implicit verbal-physical harassment. Female faculty subjectsalso viewed the incidents as more negative or problematic on several measuresthan did male faculty.

Numerous studies exist on sexual harassment in academia. Almost all ofthis research, however, deals with the sexual harassment of students by fac­ulty or with the harassment of faculty by other faculty (e.g., Adams, Kottke,& Padgitt, 1983; Benson & Thomson, 1982; Mazer & Percival, 1989; McKinney,199Oa,b; McKinney & Maroules, 1991; McKinney, Olson, & Satterfield, 1988;

IThe author would like to thank James Sikora for his assistance in collecting the data, andSusan Sprecher and Elizabeth Grauerholz for their comments on earlier versions of thispaper. This research was funded by a University Research Grant from the Office of Researchand Sponsored Programs at Illinois State University.

627

03~25/9Z112()(J-0027$06.5010 0 1992 Plenum Publishing Corporation

Page 2: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

628 McKinney

Rubin & Borgers, 1990; Schneider, 1987). The assumption has been thatsexual harassment occurs only when the offender has formal power orstatus over the victim. This assumption has'led most researchers to ignorethe possibility that students can sexually harass faculty.

Recently, a few researchers (Benson, 1984; Carroll & Ellis, 1989;Fitzgerald, Weitzman, Gold, & Omerod, 1988; Grauerholz, 1989, 1991;McKinney, 1990a; McKinney & Crittenden,- 1992; McKinney & Maroules,1991) have pointed out that the harassment'of faculty by students can anddoes occur. Benson (1984) has referred to this phenomenon as a form of"contrapower harassment." The purpose of this study was to conduct a per­son perception experiment assessing faculty perceptions of hypothetical in­cidents of sexual harassment where the offender is a 'male or femalestudent, the victim is an opposite-sex faculty member, and different typesof potentially offending behaviors are exhibited.

A variety of definitions of sexual harassment in academia exist(Crocker, 1983; McKinney & Maroules, 1991; Rubin & Borgers, 1990).These definitions tend to share several key aspects including the of­fender has power over the victim, the behavior is unwanted (as per­ceived by the victim), there is some negative harm or outcome such asdistress or interference with activities (usually as perceived by the vic­tim), and many different types of behaviors can be included as long asthere is a sexual component. In this study, one of the issues addressedwas how the type of behavior affects what is defined by subjects assexual harassment.

One useful theoretical framework for the analysis of contrapowersexual harassment is role theory (e.g., Heiss, 1981). From this perspective,the individual's ascribed and achieved statuses, and the roles associatedwith those statuses, influence perceptions and behaviors. The roles are aresult of prior socialization, past experience, and the negotiation of themeaning of certain statuses in everyday interaction. Roles may includenorms, behavioral expectations, and differing levels of power for eachstatus.

For example, individuals involved in student-faculty sexual har­assment occupy several statuses including sex, academic rank (i.e., stu­dent or 'faculty), and victim or offender. People interpret interactionsin terms of the statuses occupied and the roles played. Therefore, theperceptions of sexual harassment incidents will depend, in part, on thestatuses and roles of the individuals portrayed and of the individualsmaking judgments.

More specifically, the sex of the offender and the sex of the perceiverare expected to influence the definitions of what constitutes sexual harass­ment becallse of the sex roles associated with the status of male or female.

Page 3: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Contrapower Sexual Harassment 629

Due to greater experience with harassment and less power based on gender,it is expected that women subjects, compared to men, will indicate a greatercertainty that the incident is harassment. In addition, all subjects will indi­cate a greater certainty that the incident is harassment when the victim isfemale as opposed to male. Knowledge about the roles associated with thestatuses of victim and offender (e.g., victims are passive and less in controlthan offenders) may result in subjects judging the Victim as less responsiblethan the offender in all situations. .

A concept related to roles is that of the script (e.g., Gagnon, 1990).Essentially, scripts are cognitive representations that provide people withinformation about the appropriate who, what, where, and when in certainsituations. Scripts are acquired through socialization but are also negoti­ated, within limits, during interaction. Faculty and students will have scriptsthat govern faculty-student interactions. Certain behaviors (e.g., studentsseek academic assistance from faculty) wiII be viewed as part of the faculty­student script; other behaviors (e.g., students offer a sexual bribe to a fac­ulty member) will not. The content of these scripts may affect how behaviorand individuals in faculty-student sexual harassment incidents are judged.That is, behaviors not in the script will be seen as less appropriate or moredeviant.

LITERATURE REVIEW

There are three areas of prior research relevant to this study: (1)survey studies on contrapower sexual harassment, (2) experimental designsthat include scenarios involving student offenders, and (3) experimental de­signs (similar to that used here) focusing on academic sexual harassmentother than contrapower sexual harassment.

First, there are several studies explicitly focusing on contrapowerharassment that utilized the survey methodology. This research indicatesthat both male and female faculty report having experienced a variety ofbehaviors from their students that they define as sexual harassment. Any­where from 6% to 50% report such experiences. The exact percentagedepends on several factors including the sex of the respondent, sample,and type of measures used. Behaviors reported include sexist verbal com­ments, sexual comments on written course evaluations, obscene phonecalls believed to be from students, physical advances, uninvited sexuallooks or body language, sexual bribery, and sexual assaults (Carroll &Ellis, 1989; Fitzgerald et aI., 1988; Grauerholz, 1989, 1991; McKinney,1990a).

Page 4: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

630 McKinney

In addition, as predicted by role and script frameworks, sex differ­ences are found in this survey research. Women generally report morenegative and less tolerant attitudes toward contrapower harassment, moreexperience with such harassment, and more adverse effects of the harass­ment than men. These studies also indicate that contrapower harassmentis rarely reported to formal authorities (Carroll & Ellis, 1989; Fitzgeraldet aI., 1988; Grauerholz, 1989, 1991; McKinney, 1990a).

A second area of related research fncludes two previous experimentsthat provide information on perceptions of hypothetical incidents of har­assment by students toward other students. Pryor (1985) and Pryor andDay (1988) used college students to assess the relevance of attribution the­ory to judgments of sexual harassment. In both studies, comments depictedas having been made by a faculty member were seen as more harassingthan the same comments made by a student. This finding illustrates theimportance of the achieved role of academic rank in perceptions of sexualharassment. According to Pryor and Day (1988), whether the subject wastold to view the incident from the actor or target's point of view was alsoshown to affect judgments. That is, subjects were less likely to view theincident as harassment when they were taking the role of the offender thanwhen they were taking the role of the victim.

The final area of previous research related to this study is a groupof experiments on forms of sexual harassment in academia other than con­trapower or peer harassment. These experiments provide insight in terms offactors that affect perceptions of harassment in general. Reilly, Carpenter,Dull, and Bartlett (1982), Weber-Burdin and Rossi (1982), and Rossi andWeber-Burdin (1983) utilized factorial surveys to assess factors affectingjudgments of situations as sexual harassment. Respondents (several hun­dred undergraduates in all three studies and a small number of faculty inthe latter two studies) each read numerous vignettes describing an inter­action between a female student and a male faculty member while severalvariables (e.g., status, types of behaviors, prior relationship) were manipu­lated. Factors associated with higher ratings of the situations as harassmentwere verbal threats or physical actions by the male instructor. Factors as­sociated with lower ratings of the situations as harassment were suggestivebehaviors by the female student. In all three studies, behaviors by bothstudent and faculty member were important in affecting judgments of si­tuations as sexual harassment.

In a more recent experiment (Baker, Terpstra, & Larntz, 1990), sub­jects' reactions to harassment incidents were most strongly influenced bythe severity of the incident. Reactions were also affected by subject vari­ables including gender and attitudes toward women. Similarly, Valentine­French and Radtke (1989) found that subject gender and traditional

Page 5: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Contrapower Sexual Harassment 631

attitudes influenced attributions of responsibility in sexual harassment in­cidents. These two factors, gender and traditional attitudes, are both relatedto sex roles and gender components of scripts. '

In summary, extant research shows that a variety of forms of sexualharassment including contrapower harassment do occur in the university en­vironment and are most often not reported to formal agents of social control.Women are more concerned about, have more negative views toward, andbroader definitions of sexual harassment than men. Experimental researchindicates that factors such as subject characteristics, the type of behavior por­trayed, the situation, and the prior relationship of the individuals involvedaffect subjects' perceptions of incidents as sexual· harassment.

The empirical literature on academic sexual harassment underempha­sizes several important issues. First, faculty perceptions of these incidentshave been insufficiently studied; respondents are usually students. Second,perceptions of incidents involving student offenders have usually been ne­glected. Third, probability samples are only occasionally used. Fourth, thetype of behavior manipulated in the experiments has generally not includedforms of anonymous harassment that may be used by students (e.g., obscenecalls and comments on evaluations). Finally, measures of sexual harassmentin the experimental designs have sometimes been limited to one overalljudgment of the degree of certainty that the incident is harassment andfew additional dependent measures have been used.

The design of the present experiment incorporated features to dealwith these neglected issues. A probability sample of university faculty attwo institutions made several types of judgments after reading a hypotheti­cal situation where type of behavior (including anonymous harassment) aswell as sex of the offender were manipulated. In all instances, the offenderis the student and the victim is an opposite-sex faculty member. '

Based on the concepts of roles and scripts, and the prior literature,several hypotheses were posed. Assuming the ascribed status of gender issalient and powerful, it was predicted that subjects in the male student(female faculty) conditions compared to the subjects in the female student(male faculty) conditions would be more likely to (1) view the incident asharassment and as intended, (2) feel the student should be reported andpunished, and (3) view the student as more responsible. In addition (4)subjects will view women faculty as more upset by the incident than menfaculty. Based on people's knowledge of the roles of victim and offender(5) the offender, regardless of sex or behavior type, will be seen as moreresponsible for the incident than will the victim.

Considering role theory and the previous empirical support on sub­ject or respondent sex differences, it is hypothesized that in comparisonto male subjects, female subjects would be more likely to (6) define the

Page 6: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

632 McKinney

incidents as sexual harassment, (7) believe the faculty member should bemore upset, and (8) see the student offender as more responsible. Dueto their status and role as a faculty member, however, (9) no gender dif­ference is expected on to what extent the student offender should be re­ported or punished.

Based on the prior research that supports the importance of type ofbehavior affecting perceptions of harassment, it is predicted that (10) subjectswill be more likely to view direct, explicit.verbal and physical behavior assexual harassment, and as intentional, compared to written comments, ob­scene phone calls, and implicit behavior, which are anonymous or ambiguous.Finally, based on the assumption that subjects' scripts of faculty-student in­teractions will be unlikely to include explicit behavior, sexual comments oncourse evaluations, and obscene phone calls, it is proposed that (11) subjectswill rate these three behaviors as more definitely sexual harassment comparedto the implicit behavior condition (which might be ambiguous or subtleenough to be a part of the script). Other main and interaction effects of theindependent variables and subject sex will be explored.

METHODS

Subjects

Subjects for this experiment were faculty at two universities, IllinoisState University (ISU) and Illinois Wesleyan University (IWU). This popu­lation was accessible to the investigator, included a more diverse samplethan that from one institution, and allowed for comparisons between theperceptions of faculty at a large, public institution and a small, private in­stitution. A 50% probability sample of ISU faculty and the whole popula­tion of IWU faculty were contacted. Response rates were 52% (271 outof 525) for ISU faculty and 80% (104 out of 130) for IWU faculty. Giventhe topic of the study and the mailed questionnaire design, these responserates were considered acceptable and typical. The difference in responserates between the two institutions was probably due to the fact that allIWU subjects knew, personally, the contact person at that institution andmay have felt a greater obligation to respond.

The pool of 375 responding subjects included 151 (40%) women and217 (58%) men and 7 (2%) who did not indicate their sex. The mean ageof the subjects was 44 with a range of 24 to 67 years. The vast majority ofthe subjects were white. In terms of the population, the sample slightlyunderrepresents males but is representative for mean age. Twenty-four per­cent of the subjects (12% of the men and 12% of the women) said they

Page 7: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

contrapower Sexual Harassment 633

had personally experienced an incident similar to that described in the sce­nario. Because of the controversial topic of the study and the promise ofanonymity, no other subject information was obtained. Due to missing dataon one or more variables, full analyses were run on 313 subjects.

Procedure

Address labels were obtained through the institutional research of­fices at each university. Subjects were mailed, through campus mail, a coverletter and detailed instructions, the stimulus materials, and the dependentmeasures. The cover letter included an explanation of the project as wellas information about the voluntary and anonymous nature of the data. Sub­jects were asked to complete the questions on their own and return thematerials via campus mail to the investigator at ISU or the contact personat IWU. Reminder letters were sent to all subjects two weeks later. Afterresponses tapered off and the analyses began, a debriefing letter was sentto all department chairs to be posted or circulated to faculty members.

Design

The design of the experiment was a 2 (female student or male stu­dent) by 4 (obscene phone call or sexual comments on course evaluationsor implicit verbal and physical activity, or explicit verbal and physical ac­tivity) full factorial with random assignment of subjects to one of the eightvignettes. Subject sex was also included in the analyses.

Manipulations and Dependent Measures

The manipulations and measures described below were written bydrawing from prior experimental research, pretested on a small group offaculty, and then rewritten based on that feedback. The independent vari­ables in this experiment were sex of the student offender, type of behaviorexhibited, and subject sex. The manipulations were embedded in thevignette read by the subjects. Additional information, held constant acrossall conditions, was included to add depth and believability to the story aswell as to disguise the variables of interest. All vignettes involved an op­posite-sex interaction. The general version of the vignette follows.

Professor Jones is married and 35 years old. She (He) teaches at a medium sized,public institution in the midwest. She (He) teaches undergraduate courses in a largedepartment on campus. Last semester, Professor Jones had a male (female) studentwho was 22 years old, a senior, average in appearance, and a "B" student in one

Page 8: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

634 McKinney

of her (his) classes. This student often came to sec Professor Jones to discuss aspectsrelated to the class. One day during office hours, Professor Jones received anobscene phone call and recognized the voice of the caller to be that of this student(or When reading course evaluations at the end of the semester, Professor Jonesdiscovered one filled with explicit sexual comments and recognized the handwritingas that of this student or One day during office hours, this student sat in a ratherseductive position close to the Professor, and made comments about how attractiveProfessor Jones was or One day during office hours, this student closed the door,sat next to Professor Jones with hisiher hand on Professor. Jones' knee and askedif Professor Jones was interested in a sexual encounter with herlhim.)

Mter reading the vignette, subjects were asked to answer a series ofquestions using 7-point Likert-type continuums. These questions constitutethe dependent variables. Questions included the following: (1) subjects'perceptions of whether the incident described constituted sexual harass­ment (1: definitely is not; 7: definitely is), (2) to what extent the studentintended the behavior (1: did not, 7: did), (3) to what extent the facultymember would be upset by the incident (1: not at all; 7: extremely), (4) howmuch responsibility for the incident belongs to the student (1: not at allresponsible; 7: completely responsible), (5) how much responsibility for theincident belongs to the faculty member (same responses as Item 4),(6)whether the faculty member should report the student and hislher behaviorto the appropriate disciplinary body (1: definitely not report; 7: definitely re­port), (7) to what degree the student should be sanctioned for hislher be­havior (1: no punishment; 7: severe punishment), and (8) how common theybelieve similar situations occur at their institution (1: not at all common;7: very common). For each dependent measure, therefore, larger meansindicate greater likelihood, more certainty or greater amounts.

Finally, several demographic and background questions were askedto assess the characteristics of the subjects. These questions included sub­ject sex, age, institution, and whether he/she, personally, had ever experi­enced a situation similar to that depicted in the scenario.

RESULTS

Overview of the Analyses

The t-tests were run between the ISU and IWU faculty on each ofthe dependent variables. Results indicated no significant differences be­tween the two groups of faculty; therefore, all analyses were conducted onthe combined group.

The t-tests were also run on all of the dependent measures comparingthe means between those who reported they had personally experiencedan incident similar to the one described and those who said they had no

Page 9: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Contrapower Sexual Harassment 635

such experience. There was a significant difference on only one dependentmeasure. Subjects who reported having experienced contrapower sexualharassment indicated it was significantly more common (M = 3.7) than didsubjects who said they had no such experiences (M = 2.8; t = 4.6;P < .001).

Each dependent variable was analyzed using a 2 (offender sex) by 4(type of behavior) by 2 (subject sex) analysis of v!1riance (ANOYA). Posthoc Scheffe tests were used to test the significante between specific cellmeans. Multivariate analysis of variance was not used because the depen­dent variables were viewed as conceptually distinct and intercorrelationswere low to moderate (.1 to .6). To be more conservative, given the samplesize, only results significant at p < .01 are discussed. Cell ,sizes were morethan adequate for the analyses ranging from 76 to 180 for main effects, 29to 91 for two-way interactions, and 11 to 28 for three-way interactions. (Thesmaller cell sizes in each range occurred when the variable with four con­ditions was involved.) Table I includes the F values for all the analyses.No three-way interactions were statistically significant.

Main Effects

All three independent variables had significant main effects on sev­eral of the dependent measures. First, the sex of the student offender hadsignificant main effects for the subjects' perceptions on five dependentmeasures. More specifically, subjects in the conditions with a male studentharassing a female faculty member, compared to subjects in the conditionswith the female student/male faculty member, were significantly morelikely to view the incident as harassment (Ms = 4.81, 4.25), to feel thefaculty member would be more upset (Ms = 5.19, 4.51), to think the stu­dent was more responsible (Ms = 6.33, 5.96), to feel more strongly thatthe incident should be reported (Ms = 4.03, 3.07), and to believe that thestudent be punished (Ms = 3.72, 2.78). All of these results are in supportof the hypotheses. There were no significant effects of offender sex onperceptions of intent, faculty responsibility or the commonality of suchincidents.

In addition, the type of behavior in the incident influenced the per­ceptions of all of the dependent variables except how common subjects sawsuch incidents. Table II contains the cell means for these results and thepost hoc Scheffe tests that assess the differences between the means in thedifferent conditions. The results vary slightly by dependent measure; how­ever, in partial support of the hypotheses, the findings generally indicatethat subjects in the obscene phone call condition and the explicit verbal

Page 10: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Table I. ANOVA Results: F Values

Dependent variables

Student FacultyIndependent variables Definition Intend Upset responsible responsible Report Punish Common

Main effectsSUbject sex 6.ff' 0.2 10.3° 8.6° 0.1 5.1 3.4 3.1

Target sex 6.ff' 5.5 14.5° 6.4° 4.0 16.9" 22.4° 0.0

Behavior Type 7.ff' 8.9" 20.5° 9.5° 6.5° 19.6° 19.9" 3.0

Two-way Interactions:SUbject Sex and Target S.-:x 5.6 0.2 11.6° 0.2 1.0 0.1 ' 0.5 0.5

SUbject Sex and Behavior Type 0.7 1.7 2.1 0.6 1.1 1.9 2.7 0.2

Target Sex and Behavior Type 4.1° 0.8 4.4° 0.7 0.4 0.7 0.5 1.2

Three-way Interaction:SUbject Sex and Target Sex and Behavior Type 0.3 0.4 0.5 0.6 1.5 1.0 0.7 1.3

0p S .01.

e:'"

s::n~;-:I

~

Page 11: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Table II. Cell Means for the ANOVA Results- Mian Effects of Behavior Type on Dependent Variables with Significant F ValuesG

Dependent variables

g::I

~

1....til..~ce:.......III

;;l'"a..~

Student FacultyBehavior type Definition Intend Upset responsible responsible

Course evaluations 4.35.bc 4.79. 4.77. 6.25. 1.73.

Obscene phone calls 5.11. 5.55,0 5.37. 6.58. 1.74.

Implicit verbal and physical 3.87b 5.08. 3.85b 5.60b 2.49b

Explicit verbal and physical 4.85'0 6.01bc 5.46. 6.l8.b 2.28.b

a Means with different subscripts are significantly different at the p ~ .01 level (Sheffe).

Report

3.08.

4.54b

2.44.

4.17b

Punish

2.77.

4.23b

2.39.

3.66b

~....

Page 12: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

638 McKinney

and physical behavior condition, compared to those in the written evalu­ation and implicit behavior conditions, were more certain that the behaviorwas harassment, felt more strongly that the student intended the behavior,thought the faculty member would be more upset, and felt more stronglythat the incident should be reported and the student punished. Subjects inthe implicit behavior condition assigned the least responsibility to the stu­dent and the most responsibility to the faculty member.

Subject sex had significant main effects on three of the dependentmeasures: the certainty the behavior .was sexual harassment, the degree towhich the faculty member would be upset, and the degree of student re­sponsibility. As predicted, female subjects, compared to male subjects, weresignificantly more likely to view the incident as sexual harassment(Ms = 4.83, 4.32), to think that the faculty member should be more upset(Ms = 5.15, 4.64), and to believe that the student is more responsible(Ms = 6.38, 5.98). As hypothesized, women faculty subjects did not feelmore strongly than men that the student should be reported or punished(Ms for report = 3.35, 3.83 and for punish = 3.12, 3.46 for males andfemales, respectively).

Finally, as predicted and regardless of sex or behavior type, offenders(students) were seen as significantly more responsible than were victims(faculty; Ms = 6.2, 2.1, respectively).

Two-Way Interactions

Three two-way interactions were statistically significant. The interac­tion between the sex of the student and the type of behavior was significantfor the certainty of the behavior as harassment and for the degree to whichthe faculty member would be upset (see Table III for the cell means). First,subjects were more certain that the incident was harassment of a femalefaculty member if the behavior was course evaluations, an obscene phonecall, or implicit verbal and physical activity rather than explicit behavior.On the other hand, they were more certain the scenario involved harass­ment of a male faculty member if the behavior was explicit verbal andphysical activity compared to the other three behavior types.

Second, subjects believed the female faculty member would be moreupset than the male faculty member in the course evaluation, obscenephone call, and implicit activity conditions, but equally upset as the malefaculty member in the explicit activity condition.

The interaction between sex of the student and sex of the subject wassignificant for the degree to which the faculty member would be upset.More specifically, female subjects believed that male and female faculty

Page 13: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Table III. Cell Means for Interaction Between Target Sex and Behavior Type for Delinition and Upset (DelinlUpset) (p s .01)

Behavior type

Qaiil'g:Ii.....til~=a--..iil'"S..a

Course evaluations

Target sex

Male professor and female student

Delin

3.61

Upset

4.03

Obscene calls Implicit behavior

Delin Upset Delin Upset

4.71 4.91 3.75 3.70

Explicit behavior

Delin Upset

4.97 5.44

Female professor and male student 5.00 5.42 5.44 5.76 4.00 4.00 4.72 5.49

e:IC

Page 14: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

640 McKinney

would be equally upset (Ms = 5:11 for male faculty and 5.18 for femalefaculty) whereas male subjects reported that the female faculty memberwould be more upset than the male faculty member (Ms = 4.10 for themale faculty member and 5.19 for the female faculty member).

DISCUSSION

Clearly, one concise interpretation of the results of this experimentis, as predicted, that the status and the role expectations of gender had apowerful influence on the subjects' perceptions of contrapower sexual har­assment incidents. Both subject sex and victim/offender sex influenced thesejudgments. Faculty member and offender roles also influenced perceptions.Consistent with prior research, the type of behavior portrayed affected per­ceptions of the incident.

Role theory, and the concept of sex roles in particular, suggests thepotential salience of gender on perceptions of sexua(harassment. As statedin the introduction, survey research on sexual harassment supports genderdifferences in attitudes about sexual harassment. This study provided em­pirical support for the importance of both perceiver (subject) sex and actor(offender/victim) sex on perceptions of contrapower harassment.

This study validated survey research on sex differences (Carroll &Ellis, 1989; Fitzgerald et aI., 1988; McKinney, 1990a) by showing that fe­male subjects were more likely to see the incidents as harassment, believethe victim would be upset, and indicate that the student was responsibleas compared to male subjects.

In addition, the results indicated a traditional and perhaps paternalor protective attitude toward female faculty as victims. Male and femalesubjects saw the incidents as more problematic and deserving of moreattention when the victim was female. Male subjects also thought thatthe female faculty member would be more upset than the male facultymember, whereas female subjects thought male and female faculty wouldbe equally upset. These findings also supported survey results on actualincidents of contrapower harassment showing that women faculty do re­port greater negative perceptions of and reactions toward contrapowerharassment incidents than male faculty (Carroll & Ellis, 1989; McKinney,1990a).

The relevance of the concept of roles is seen, not only in the findingsabout gender, but also in other results. As predicted, the role of a facultymember may be more salient than that of gender when it comes to makingjudgments about reporting or punishing the students. The sex of the facultysubject had no effect on these judgments. It is likely that faculty subjects,

Page 15: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

contrapower Sexual Harassment 641

in role playing the situation of the faculty victim, use their knowledge ofthe roles of faculty and student. They may believe that, as faculty, theyhave the power, ability, or responsibility to handle the situation similarlywhether they are male or female.

Finally, beliefs about what it means to play the victim and the of­fender role were seen in the result that offenders were judged more re­sponsible for the incident regardless of their se~ or the specific type ofbehavior in which they engaged. This finding is in line with the researchby Pryor and Day (1988) showing the importance of perspective on per­ceptions. That is, given that all the subjects were university faculty, theywere probably viewing the incident from the faculty/victim perspective and,therefore, judging the student as more responsible than the faculty memberin the vignette. This finding is also a form of a manipulation check showingthat subjects did accurately read and perceive that the student engaged inthe harassing behavior.

In support of earlier research on forms of sexual harassment otherthan contrapower sexual harassment (e.g., Baker et aI., 1990; Reilly et aI.,1982; Rossi & Weber-Burdin, 1983; Weber-Burdin & Rossi, 1982), type ofbehavior portrayed in the vignette affected perceptions by the subjects. Ingeneral, obscene phone calls and explicit verbal/physical behavior were per­ceived as most serious and negative by subjects. Perhaps subjects believedthese behaviors were more difficult to explain away as a joke or as mis­communication compared to comments on evaluations or implicit behavior.In fact, subjects were more certain that the student intended these formertwo behaviors compared to written comments and implicit behavior. It isalso likely that subjects' scripts for faculty-student interactions did not in­clude these types of behavior so that the behaviors were viewed as inap­propriate or nonnormative for that relationship.

Some interesting interactions occurred between victim gender andtype of behavior. Only in the condition of explicit verbal/physical behaviordid subjects feel male faculty would be as upset as female faculty and weresubjects more certain that the incident was harassment. Subjects, then,seem uncertain about labeling sexual behaviors by a female student towarda male faculty member as harassment and negative. Such labeling was easi­est for them to do in the least ambiguous condition of explicit verbal/physi­cal behavior. Perhaps male and female faculty members were less clearabout when sexual behaviors by a female student directed toward a malefaculty are harassment because they believe the male faculty member (moreso than a female faculty member) desired or encouraged the behaviors, orwill be unaffected (not harassed) by the behaviors.

This experiment, of course, has limitations. Subjects were asked torole play or respond to hypothetical incidents. Results might have been

Page 16: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

642 McKinney

different had subjects been asked about their own actual experiences orif observational methods were used. In addition, victim and offender sexwere confounded in this design be~ause all incidents were opposite-sexincidents. Future studies should orthogonally manipulate these variablesto ascertain whether results are due to victim or offender sex or an in­teraction of those two variables. This latter strategy, however, would in­troduce the complexity of including both heterosexual and homosexualharassment incidents. . '

Future experimental research on contrapower harassment should alsoinvolve the manipulation of other independent variables such as prior stu­dent-faculty relationship and response of the victim. In addition, the per­ceptions of both students and administrators as subjects should be assessed,along with those of faculty members, to compare the effects of differentacademic statuses and roles on perceptions of sexual harassment. Assessingthe actual content of subjects' scripts for faculty-student interactions wouldallow for testing the importance of such scripts on perceptions of harass­ment. This might be accomplished with open-ended questions on a surveyor an interview. Finally, subject characteristics, such as sex role, need tobe added to the designs.

The findings of this study have policy implications for university ad­ministrators. For example, administrators who write policies, plan educa­tional programs, or hear harassment complaints should be aware thatwomen are more likely to label contrapower incidents as sexual harassmentand to see them as more problematic than men. The results also indicatethat both male and female faculty members prefer or believe they shouldhandle contrapower incidents rather than formally reporting or punishingthe students. Administrators need to provide information and education toassist faculty members in handling these incidents informally.

This research also provides some information on what type of behav­ior is viewed, by faculty, as most problematic. Administrators should con­sider including obscene phone calls and explicit verbal/physical sexualbehaviors as types of contrapower sexual harassment in written policy state­ments and educational materials. Finally, administrators dealing with sexualharassment should take into account that certain forms of harassment areseen as more ambiguous than others and that cases involving ambiguousforms may be more difficult to resolve.

REFERENCES

Adams, J. W., Kollke, H. L. & Padgill, J. S. (1983, November). Sexual harassment of universitystudents. JOllma[ of College Personnel, 484-491.

Page 17: Contrapower sexual harassment: The effects of student sex and type of behavior on faculty perceptions

Contrapower Sexual Harassment 643

Baker, D. D., Terpstra, D. E., & Larntz, K. (1990). The influence of individual characteristicsand severity of harassing behavior on reactions to sexual harassment. Sex Roles, 22,305-325.

Benson, D. J., & Thomson, G. E. (1982). Sexual harassment on a university campus: Theconfluence of authority relations, sexual interest and gender stratification. Social Problems,29, 236-251.

Benson, K. (1984). Comment on Crocker's 'an analysis of university definitions of sexualharassment.' Signs, 9, 516-519.

Carroll, L., & Ellis, K. L. (1989). Faculty attitudes toward sexual harassment: Survey results,survey process. Initiath'es, 52, 35-41. ,

Crocker, P. L. (1983). An analysis of university definitions.'of sexual harassment. Signs, 8,696·707. .

Fitzgerald, L. F., Weitzman, L. M., Gold, Y., & Ormerod, M. (1988). Academic harassment:Sex and denial in scholarly garb. Psychology of Women Quarterly, 12, 329·340.

Gagnon, J. (1990). The explicit and implicit use of the scripting perspective in sex research.In J. Bancroft (Ed.), Annual re~'ielV of sex research. Society for the Scientific Study of Sex.

Grauerholz, E. (1989). Sexual harassment of women professors by students: Exploring thedynamics of power, authority, and gender in a university setting. Sex Roles, 21, 789-801.

Grauerholz, E. (1991). Sexual harassment in the academy: The experience of faculty women.Unpublished manuscript.

Heiss, J. (1981). Social roles. In M. Rosenberg & R. H. Turner (Eds.), Social Psychology:Sociological Perspecti~'es. Basic Books: New York.

Mazer, D. B., & Percival, E. F. (1989). Students' experiences of sexual harassment at a smalluniversity. Sex Roles, 20, 1-22.

McKinney, K. (1990a). Sexual harassment of university faculty by colleagues and students. SexRoles, 23, 421-43.

McKinney, K. (1990b). Attitudes toward sexual harassment and perceptions of blame: Viewsof male and female graduate students. Free Inquiry into Creath'e Sociology, 18, 191-196.

McKinney, K" & Crittenden, K. (1992). Contrapower sexual harassment: The offenders'viewpoint. Free Inquiry into Creatire Sociology, 20, 3-20.

McKinney, K., & Maroules, N. (1991). Sexual Harassment. In E. Grauerholz & M. Koralewski(Eds.) Sexual coercion. Lexington, MA: Lexington Press.

McKinney, K., Olson, C., & Satterfield, A. (1988). Graduate students' experiences with andresponses to sexual harassment. Journal of Interpersonal Violence, 3, 319-325.

Pryor, J. (1985). The lay person's understanding of sexual harassment. Sex Roles, 13, 272-286.Pryor, J., & Day, J. (1988). Interpretations of sexual harassment: An attributional analysis.

Sex Roles, 18, 405·417.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, 99·110.Rossi, P. H., & Weber-Burdin, E. (1983). Sexual harassment on the campus. Social Science

Research, 12, 131-158.Rubin, L. J., & Borgers, S. B. (1990). Sexual harassment in the universities during the 1980s.

Sex Roles, 23, 397,411.Schneider, B. E. (1987). Graduate women, sexual harassment and university policy. Journal

of Higher Education, 38, 46-65.Valentine-French, S., & Rudtke, H. L. (1989). Attributions of responsibility for an incident

of sexual harassment in a university setting. Sex Roles, 21, 545-555.Weber-Burdin, E., & Rossi, P. H. (1982). Defining sexual harassment on campus: A replication

and extension. Journal of Social Issues, 38, 111-120.