Você está na página 1de 14

Factors Affecting Perceptions of Workplace Sexual Harassment

Author(s): Thomas W. Dougherty, Daniel B. Turban, Diane England Olson, Peggy D. Dwyer,
Melody W. Lapreze
Source: Journal of Organizational Behavior, Vol. 17, No. 5 (Sep., 1996), pp. 489-501
Published by: John Wiley & Sons
Stable URL: http://www.jstor.org/stable/2488557
Accessed: 25/07/2009 08:53

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=jwiley.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

John Wiley & Sons is collaborating with JSTOR to digitize, preserve and extend access to Journal of
Organizational Behavior.

http://www.jstor.org
JOURNAL OF ORGANIZATIONAL BEHAVIOR, VOL. 17, 489-501 (1996)

Research Factors affecting perceptionsof workplace


Note sexual harassment

THOMAS W. DOUGHERTY AND DANIEL B. TURBAN


Department of Management, Universityof Missouri, Columbia, MO 65211, U.S.A.

DIANE ENGLAND OLSON


Department of OrganizationalBehavior & Environment,California State University-Sacramento,
Sacramento, CA 95819, U.S.A.

PEGGY D. DWYER
School of Accountancy, Southern Illinois University, Carbondale,IL 62901, U.S.A.

MELODY W. LAPREZE
Stephens College, Columbia, MO 65211, U.S.A.

Summary This study investigated several variables that determine how one interprets another's
behavior as sexually harassing in ambiguous situations. Data were collected from
1234 male and female graduate and undergraduate student subjects who responded
to 24 versions of a vignette describing an interaction between a male and female who
worked for the same firm (2 x 3 x 2 x 2 design). Variables manipulated included
(1) amount of prior socializing of the man and woman; (2) job status (peer coworker,
direct supervisor, executive) of harasser; (3) verbal versus touching behavior; and
(4) physical setting (at computer terminal or happy hour). Results revealed several
main effects such that a male's potentially harassing behavior toward a female was
evaluated more negatively (1) when the pair had not previously socialized, (2) when
the harasser was a supervisor or executive, (3) when it involved verbal comments, and
(4) when the behavior occurred in a work (versus social) setting. Additional analysis
revealed that, contrary to previous studies, there were no male-female differences
among respondents in evaluations of behavior as harassing.

Introduction

Sexual harassment in the workplace is a problem especially for women, in that harassment can
damage women's prospects for gaining employment, advancement, and wages. In addition,
harassment can create an offensive, hostile, and intimidating work environment that interferes
with one's performance and job success. The prevailing definitions of sexual harassment,
including the 1980 EEOC Guidelines, suggest the importance of the recipient of harassment's
evaluation of sexual behavior received at work (Konrad and Gutek, 1986). Although some
behaviors are direct and extreme and generally regarded as sexual harassment, other behaviors
(e.g. personal comments) are more ambiguous and seen as harassing by some but not by others.
The hearings concerning the alleged sexual harassment of Anita Hill by U.S. Supreme Court

CCC 0894-3796/96/050489-13 Received 11 February 1994


(D 1996 by John Wiley & Sons, Ltd. Accepted 23 August 1995
490 T. W. DOUGHERTY ET AL.

nominee Clarence Thomas in 1991 dramatized the ambiguities inherent in interpretation of work
behavior as harassing. The present research considers contextual conditions relevant to these
interpretations, providing clarity about several personal and situational characteristics that can
determine one's evaluation of ambiguous job behavior as sexual harassment.
Since the 1980s, a number of studies of sexual harassment have appeared, most of them
descriptive in nature. Many of these studies have been surveys focusing on the frequency of
harassment, who are the perpetrators, and who are the recipients (see Fitzgerald and Shullman,
1993; Gutek, 1985; USMSPB, 1981; Tangri, Burt and Johnson, 1982). For example, Gutek
(1985) reported that male harassers tend to be older, married, and supervisors, while female
harassers tend to be young, unmarried, attractive, and in nonsupervisory positions. Female
victims tend to be single, while married or widowed women have fewer harassment experiences.
Although most sexual harassment research has been somewhat descriptive and exploratory,
some scholars, including Gutek (1985) and Tangri et al. (1982), and Fitzgerald, Hulin and
Drasgow (1992) have offered explanatory models. The arguably most useful model is the 'sex-
role spillover model' (Gutek, 1985), an outgrowth of role theory, which focuses on the various
sets of expectations associated with different kinds of roles (Katz and Kahn, 1978). The sex-role
spillover model includes the notion that when a job comes to be seen as a man's or woman's job,
the appropriate gender role 'spills over' to become formally or informally incorporated into the
work role. Gender cues became more salient, such that men and women may come to be treated
not as colleagues but according to male or female stereotypes.
Our research acknowledges the need for more attention to be given to 'context' in studying
behavior in organizations (Jackson and Schuler, 1995) by focusing on how situational variables
may influence perceptions of ambiguous behavior that may be perceived as sexually harassing.
Role theory, which provides the theoretical underpinnings of the sex-role spillover model
(Gutek, 1985), provides a theoretical framework for our research. Role theory proposes that
people hold differing expectations for their own and others' behavior according to roles they
fulfill (Katz and Kahn, 1978). This role behavior perspective provides a framework for our
hypotheses, which predict differences in reactions to behavior of a potentially harassing nature
according to differing contexts which influence role expectations. These differing role contexts
can stem from contextual factors such as differing job status, social versus work settings, or
friendship versus business relationships.
As part of a focus on context, to understand perceptions of sexual harassment researchers
must clarify relevant situational factors. Social-sexual situations can be ambiguous, and a variety
of contextual factors related to role expectations for a potential harasser may determine how
positively or negatively one perceives behavior. For the present study, we contribute to the
understanding of harassment by focusing on several key contextual factors associated with role
expectations that are expected to make a difference in how people interpret ambiguous social-
sexual situations at work. In addition, we use a more comprehensive set of dependent measures
compared to previous research. We now discuss the contextual factors thought to influence
perceptions of another's behavior and review the relevant literature.

Nature of the relationshipbetween the parties


Brewer (1982) pointed out that sexual harassment should be studied in the context of
interpersonal communication and dyadic processes. Little research, however, has been
conducted from this perspective. The nature of the relationship between two people in a
potentially sexually harassing situation may be important for how an ambiguous situation is
WORKPLACE SEXUAL HARASSMENT 491

interpreted. Prior relationships between the parties influence role expectations and therefore may
influence perceptions of harassment because the meaning and evaluation of behaviors are
different for those in prior relationships. A woman, for example, may be less offended by verbal
comments or even touching by a male she has socialized with because her perception of the
meaning of the behavior suggests that such behaviors are appropriate. Reilly, Carpenter, Dull,
and Bartlett (1982) examined this issue in a study of 232 undergraduate students at the
University of California, Santa Barbara, who were surveyed in 1981. The survey included
vignettes varying eight situational and behavioral factors potentially relevant to judgements of
sexual harassment of students by professors. They found that harassment was judged signifi-
cantly lower if the professor and student were friends, had gone out several times, or had been
regularly dating. Similarly, in a study using the same instrument with 59 students at the
University of Massachusetts, these findings were confirmed (Weber-Burdin and Rossi, 1982).
Thus, only a few studies have examined prior interpersonal relationships in interpretation of
social-sexual behavior as harassment. Moreover, these studies were not performed in a work
context. Research in more of a work context would be valuable from the standpoint of
generalizability of findings to work settings. We expect that past socializing between a man and a
women will be a factor related to more positive perceptions of potentially harassing behaviors at
work.

Job status of the harasser


The status of a potential harasser in the organization is likely to be related to whether a victim
perceives an ambiguous situation as harassment. Job status is related to a power differential
between two persons, which could influence perceptions of behavior (see Cleveland and Kerst,
1993). For example, supervisors, because of their role, may be held to higher standards of
behavior than coworkers (Gutek, Morasch and Cohen, 1983), in that they have power as
representatives of the organization. Thus, potentially harassing behaviors may be seen as more
offensive coming from supervisors compared to coming from peers. Similarly, behavior seen as
friendly from coworkers may be seen as a threat when coming from supervisors, because of the
power supervisors have to obtain compliance. In this vein, results of the U.S. Merit Standards
Protection Board Survey indicated that workers in the lowest status positions are those most
likely to see sexual harassment by supervisors as serious problems (Brewer, 1982).
Gutek et al. (1983) manipulated three independent variables, including job status, in a vignette
study of harassment using 218 undergraduate students. They found that ambiguous, potentially
sexual behavior initiated by a higher to a lower status person was evaluated more negatively than
when toward an equal or higher status person. In another analysis, Collins and Blodgett (1981)
received surveys from 1846 Harvard Business Review subscribers to assess perceptions of sexual
harassment. On the basis of opinions about six statements made by females, they concluded that
the perceived seriousness of a potentially harassing behavior is higher when coming from a
supervisor compared to a coworker. Bursik (1992) found that students perceived behavior from
another student as less likely to be harassment compared to a professor's behavior. This status/
power effect was stronger for less blatant behavior. Similarly, in the Reilly et al. (1982) vignette
study of undergraduates, harassing behaviors from professors tended to be more frequently
related to ratings of harassment, compared to behaviors from graduate TAs, or non-TA
graduate students.
In contrast, the replication study of Weber-Burdin and Rossi (1982) did not find a difference
depending on whether behavior was from TAs or professors. Similarly, Fitzgerald and
492 T. W. DOUGHERTY ET AL.

Ormerod's (1991) vignette study of university faculty and students found that whether or not a
professor was in direct authority over a student had little effect on perceptions of behavior as
harassing. Sheffey and Tindale's (1992) vignette study using undergraduates also found no main
effects for degree of status differential between a superior and subordinate on perceptions of the
superior's harassing behavior.
Littler-Bishop, Seidler-Feller and Opaluch (1982) supported the hypothesis that the lower the
status of the harasser (e.g. ticket agent versus pilot) the more negative were flight attendants'
feelings about moderate verbal and physical harassment. Some women, then, may be socialized
to be more tolerant of high status males. Although there may appear to be inconsistent results for
the effects of status, these conflicting results could reflect a confusion of social status and job
status accompanied by power (Brewer, 1982). Females may be more tolerant of potentially
harassing behavior from those with higher social status (e.g. pilots) but less tolerant of those in
direct power over them at work (e.g. supervisors). In light of the weight of evidence from
previous research and our earlier arguments, we expect that subjects' perceptions will be more
negative in evaluating an ambiguous situation involving potentially sexual behavior from a
person with a higher job status, that is, a person with power over the target person.

Physical versus verbal behavior

A variety of actions may be interpreted as harassing at work. In general, we might expect


touching to be viewed as more intrinsically sexual than verbal comments because of violations of
one's 'body space'. In the Gutek et al. (1983) vignette study of undergraduates, touching
behavior was more negatively evaluated than verbal comments. They also found, unexpectedly,
that touching, when combined with comments about one's body, was viewed more favorably
than touching combined with a work-related comment. The negative view of touching with work
comments was even stronger for men than women. Similarly, Marks and Nelson (1993) found
that potentially harassing behavior from professors which contained touching in addition to
verbal comments was seen by students as more inappropriate than verbal behavior only.
Conversely, Reilly et al. (1982), studying sexual behavior from instructors, found that no striking
verbal-physical behavior differences were apparent. Nevertheless, for the present study, because
violations of one's body space should produce more negative reactions, and in light of previous
research evidence, we expect that in a somewhat ambiguous situation, touching behavior will be
perceived more negatively than verbal behavior.

Physical setting

Interactions among employees of a social-sexual nature take place in a variety of settings. Some
interactions are in the primary work setting, while others are in settings of a more peripheral
nature, including a company Christmas party or an after-work 'happy hour'. It seems likely that
behaviors of a potentially sexual nature may be interpretedquite differently according to setting,
but this has been studied very little, especially in a work context. The Reilly et al. (1982) study of
student-instructor relations did examine the social setting of potentially harassing behavior.
They found that there were no differences in perceptions whether the interaction occurred in
instructors' offices, the library, a restaurant, or a party. However, we believe that typically in
university campus settings, students' social and work roles are more overlapping and the setting
WORKPLACE SEXUAL HARASSMENT 493

boundaries less clear compared to the workplace. We expect that in a social setting, sexual
behavior from an organization member may be seen as less threatening, because it is not so
clearly tied to work and career outcomes. The most negative evaluation of sexual behavior would
be likely to occur in the primary work setting.

Sex of respondents

The literature of sexual harassment consistently supports the notion that females evaluate
behavior of a social-sexual nature differently from males (e.g. Gutek, 1985; Marks and Nelson,
1993; Mazer and Percival, 1989; Kenig and Ryan, 1986). Fitzgerald and Shullman (1993) report
that probably the most stable finding in the harassment literature is that women are more likely
than men to view certain behavior as harassing, especially in less explicit situations. Males for
example, are more likely to be complimented by sexual behavior, while females are likely to be
offended (Collins and Blodgett, 1981; Gutek, 1985; Reilly et al., 1982). Similarly, Gutek (1985)
reported that 84 per cent of females in her large survey of working persons viewed sexual
touching as harassment, compared to only 58 per cent of males. Although previous studies reveal
consistent male-female differences in perceptions, it is important to obtain more current data,
especially considering the increase in awareness and the media attention given to harassment
since the 1980s. This increased awareness may have produced some narrowing of the gender
'perception gap'. However, because of the consistency of previous findings, we hypothesize that
males will show a more positive attitude than females when presented with ambiguous situations
including sexual behavior in the workplace.

Respondents'previous experience with harassment

Cohen and Gutek (1985) suggested that previous experience with harassment should relate to
how people cognitively organize and categorize the subject. One who has experienced past
harassment would be expected to be more sensitive to harassment and to interpret ambiguous
behaviors as harassment, compared to one with little or no experience.
Conversely, 'habituation' to sexual harassment could also occur, in which persons
experiencing harassment become somewhat desensitized to sexually harassing behavior.
Konrad and Gutek (1986) also found that because of habituation, workers who experienced
sexual behaviors at work often do not label the behaviors as harassment. Mazer and Percival
(1989) also found an absence of relationships of harassment experience to perceptions and
attitudes. Similarly, the Collins and Blodgett survey (1981) of Harvard Business Reviews readers
found that respondents who had knowledge of certain sexual behavior in their organizations
were more likely to assert that these behaviors are not harassment. However, it is not clear the
extent to which this knowledge of sexual behavior in the organization reflects one's personal
experience. Thus, arguments exist for both the 'heightened sensitivity' and the 'habituation'
phenomenon as they relate to previous experience with harassment. Because of the increased
awareness, information, and education about sexual harassment throughout the 1980s, we
anticipate that more sensitivity to harassment is likely for those with personal experience with
harassment. Therefore, we hypothesize that to the extent that one has personally experienced
sexual harassment, he or she will more negatively evaluate ambiguous behavior which may be
harassment.
494 T. W. DOUGHERTY ET AL.

Our research involved subjects' responding to scenarios including a male and a female, and
involving potentially harassing behavior. The study tested the following hypotheses about
perceptions of harassment in a work context:
(1) Potentially sexually harassing behavior will be perceived more positively when the
persons involved have frequently socialized in the past, compared to when they have
not socialized.
(2) Potentially sexually harassing behavior initiated by a higher level executive or direct
supervisor will be perceived more negatively than when initiated by a coworker.
(3) Potentially sexually harassing behavior will be perceived more negatively when it
involves touching, compared to when the behavior is verbal.
(4) Potentially sexually harassing behavior occurring within the primary work setting will
be perceived more negatively than when occurring in a more social setting.
(5) Male respondents will evaluate potentially sexually harassing behavior more positively
than women.
(6) Respondents who have themselves experienced more sexual harassment will evaluate
potentially sexually harassing behavior more negatively than those with less experience
with harassment.
As exploratory analyses, we also assessed the interactions among the independent variables,
for possible additional insights into perceptions of sexual harassment.

Method
Subjects
Subjects were 1234 undergraduate and graduate/professional students who returned surveys
mailed in the spring of 1990 to a random sample of this population at a large midwestern state
university (30 per cent response rate). A large majority were white (91 per cent), and there were
more female (62 per cent) than male respondents. Finally, there were similar numbers of
graduate students (27 per cent), seniors (20 per cent), juniors (18 per cent), sophomores (16 per
cent), and freshmen (19 per cent). This study was a part of a large survey initiated by a university
task force on gender sensitivity, of which the first author was a member.

Materials
Four independent variables were manipulated in a vignette about an interaction between a man
(John) and a woman (Jane) who work for the same organization. Each respondent was presented
with one of 24 versions of the vignette (2 x 2 x 2 x 3). The first manipulated variable was the
extent of prior socializing outside of work. In one level, the male and female had rarely socialized;
in the other level they had frequently socialized.
The second manipulated variable was the job status of the male (John). In one condition, he
was a peer coworker. In the second condition, he was the female's direct supervisor. In the third
condition, he was a high level executive.
The third manipulated variable was touchingversus verbal behavior. In one level the male put
his arm around the female; in the other level he verbally asked 'How is your love life, have you
had any exciting dates lately?'
WORKPLACE SEXUAL HARASSMENT 495

The fourth manipulated variable was the setting of the interaction. The female (Jane) was
either working late at the computer terminal, or sitting in a bar across the street for the after-
work happy hour.
An example scenario: Jane is a typist at ABC Insurance Company. She is working late and is
sitting at the computer terminal when she hears someone say 'How is your love life going? Have
you had any exciting dates lately?' She looks up and sees that it is John, her direct supervisor. She
has worked with John for several years but has rarely socialized with him outside of the work
setting.
The fifth variable analyzed, which was not manipulated, was the sex of the respondent, male or
female.
The sixth variable analyzed was the respondents'previousexperienceof sexual harassmentat the
University. In a separate portion of the survey, respondents were asked to report on a 5-point
scale (1 = never; 5 = frequently) the frequency of their experience of (1) unwanted sexual remarks
about clothing, bodies, or sexual activities; (2) unwanted sexually suggestive looks or gestures;
(3) unwanted flirting or sexual teasing; (4) unwanted pressure for dates or social activity;
(5) unwanted letters or phone calls which were sexually motivated; (6) unwanted deliberate
touching, kissing, leaning over, cornering or pinching; and (7) attempted rape or sexual assault.
Respondents reported their experience with these behaviors from several sources, including
professors, teaching assistants, staff, administrators, and other students. For each respondent, the
sum across the seven items for the five sources was summed for a total 'experienced harassment'
score used in the present analysis. For these 35 items, the alpha was 0.86. Since respondents were
randomly assigned to treatment conditions, we did not expect previous harassment experience to
be correlated with the independent variables. This expectation was confirmed (see Table 2).

Dependent measures
Respondents indicated their agreement (on 5-point scales) with 15 items describing reactions to
the behaviors described in the vignette. An iterated principal factor analysis with varimax
rotation and squared multiple correlations as initial commonality estimates suggested three
factors. These three factors were labelled Controlling/Power Behavior, Positive Reaction, and
Negative Reaction, and explained 34 per cent, 17 per cent, and 17 per cent of the variance,
respectively (see Table 1). Coefficient alpha for scales created with unit weights for each of the
variables were 0.92, 0.88, and 0.89, respectively.
Respondents also described John's behavior on 14 (7-point) semantic differential items. For
example, respondents indicated the extent to which John's behavior was 'hostile or not hostile',
'considerate or not considerate', etc. by placing a check in the appropriate blank between these
descriptors. Factor analyses suggested one factor, and a scale, labelled Negative Behavior was
created with unit weights of the 12 items that loaded on the one factor (alpha = 0.95). Table 2
displays correlations among all study variables, indicating that the dependent variables'
intercorrelations ranged from -0.34 to 0.73.

Results

We conducted a multivariate analysis of variance to investigate the effects of prior socializing,


harasser status, touch, setting, and respondent sex on the dependent variables. As indicated by
496 T. W. DOUGHERTYET AL.

Table 1. Factor analyses of dependent measures


Controlling/power Positive Negative
behavior reaction reaction
8. John is attempting to control Jane 85 -14 10
7. John is trying to exert power over Jane 85 - 14 12
9. John is attempting to dominate Jane 85 - 15 10
2. John's behavior creates an offensive work environment 77 - 13 25
3. John's behavior creates a hostile or intimidating work
environment 76 -10 25
4. John is making sexual advances toward Jane 72 -7 22
1. John's behavior could interfere with Jane's future
work performance 69 -4 19
6. John is physically attracted to Jane 40 24 18
5. John is just trying to be friendly -63 31 - 18
11. Jane is flattered by John's behavior - 10 89 - 17
12. Jane views John's behavior as complimentary - 14 85 - 18
10. Jane is happy to get this attention from John - 15 84 - 18
14. Jane feels embarrassed by John's behavior 21 - 13 88
13. Jane feels uncomfortable with John's behavior 23 -23 82
15. Jane feels offended by John's behavior 33 -25 80
Numbersare factorloadings.The underlinedfactorloadingswereused in scaleconstruction.

the Wilks' Criterion, there were significant multivariate main effects for touch, F(4, 997) = 34.68,
p < 0.0001, socializing F(4,997) = 18.33, p < 0.0001, status, F(8, 1994) = 3.72, p ? 0.0003, and
setting, F(4, 997) = 7.53, p < 0.0001. Exploratory analysis of interactions revealed a significant
touch by socializing interaction, F(4, 997) =2.99, p < 0.0 180. No other interactions were
significant.
Follow-up univariate results (see Table 3) indicated that the extent to which John and Jane had
previously socialized influenced all of the dependent variables. Specifically, when John and Jane
frequently socialized, John's behavior was seen as less controlling, Jane had more positive
reactions and less negative reactions, and John's behavior was seen as less negative. Eta-squared
ranged from 3 per cent to 5 per cent of the variance. Status influenced only controlling behavior.
Post-hoc analyses indicated that the supervisor and the executive were seen as more controlling
than the peer (eta-squared 1 per cent). In addition, verbal comments, contrasted with touching,
led to perceptions of greater controlling/power behavior, less positive reactions by Jane, and
perceptions that John's behavior was more negative. Eta-squared ranged from 0.3 per cent to
3 per cent of variance. The setting of the incident only influenced perceptions of controlling
behavior, such that when the incident occurred at work, John was seen as more controlling (eta-
squared 1 per cent). Finally, respondent sex did not influence any of the dependent measures.
To assess the relationship of respondents' previous harassment experience with perceptions of
harassment, we computed the Pearson correlation between the amount of harassment (summary
index) and the dependent variables, since amount of harassment was a continuous variable.
Considering the correlation of harassment experience with the four dependent variables, only the
correlation with controlling/power behavior was significant (r = 0.06, p < 0.05). We also note
that females had experienced more harassment (r = -0.23, p < 0.01).
Turning to the touch by socialization interaction, follow-up univariate results indicated a
significant (at p < 0.05) interaction only for negative reactions. Examination of this interaction
revealed that there was a large decline in negative reactions to verbal comments when John and
WORKPLACE SEXUAL HARASSMENT 497

10.9. 8. 7. 6. 5. 4. 3. 2. 1.
2 Table
Sex= Job 2.
Coefficient Prior
(1 Setting
Positive
Negative
NegativePrevious
=
alpha (0 Touching
status
in (0
executive)
the socializing Descriptive
reactionfemale, =bar,
reaction 1 (0 behavior
behaviorControlling/power
experience=
= (0
of
diagonal. =coworker, statistics
2=male)
N behavior 1 work) rarely,
sexual 1
=verbal,
= 1
ranges
=
from
=supervisor,
harassment
1080 touch)
frequently)
to

1234.47.74
9.18 20.97
8.20 1.39 1.00.50
43.14 0.49 Mean
0.50
*p
<
2.41 7.13
0.05;13.962.18 9.03
0.49 0.820.50 0.50 S.D.
0.50
tp
< 0.1 1
-0.04
-0.19tIt 0.07* -0.05
-0.02 0.03
0.00 0.06*
0.01.
2
-0.01
-0.22t -0.18t-0.02 -0.05
0.20t
-0.24t 0.05

3
-0.03 -0.01
0.02 0.08*0.05 -0.03
0.00

0. 4
0.02
0.04 lOt0.00
0.04 0.00

5
-0.03 -0.02
0.05
0.00 -0.23t

6
0.04 -0.05
-0.02 0.86
0.06*

7
-0.73t
0.52t 0.92
-0.34t

8
0.88
-0.43t
-0.53t

9
0.89
0.61t

10
0.95
498 T. W. DOUGHERTYET AL.

Table 3. Means and standard deviations of independent variables: by dependent variable


Independent Controlling/power Positive Negative Negative
variable behavior reactions reactions behavior

Touch
Verbal 21.5* (7.1) 8.0* (2.2) 9.3 (2.5) 50.1* (13.8)
Touched 20.6* (7.0) 8.4* (2.2) 9.1 (2.3) 45.0* (13.7)
Socialized
Infrequently 22.3* (7.1) 7.7* (2.3) 9.8* (2.4) 50.7* (14.4)
Frequently 19.8* (6.8) 8.7* (1.9) 8.6* (2.3) 44.7* (12.9)
Setting
Bar 20.4* (6.9) 8.2 (2.1) 9.3 (2.4) 47.3 (14.2)
Work 21.7* (7.2) 8.2 (2.2) 9.1 (2.4) 48.0 (13.8)
Status
Peer 19.8t (7. 1) 8.1 (2.3) 9.0 (2.5) 46.9 (14.2)
Supervisor 21.7* (7.4) 8.2 (2.2) 9.1 (2.4) 48.0 (14.5)
Executive 21.6* (6.6) 8.3 (2.0) 9.3 (2.2) 48.0 (13.3)

Respondent sex
Female 21.2 (7.3) 8.2 (2.3) 9.2 (2.5) 47.8 (14.4)
Male 20.8 (6.6) 8.3 (1.9) 9.2 (2.2) 47.3 (13.3)

Numbersin parenthesesare standarddeviations.


variable.
* Therewas a significant(at p < 0.05) differencein the meansbetweenthe levelsof the independent
different(at p < 0.05) fromboth supervisorand executive.
t Peerwas significantly

Jane had frequency socialized. It also appears that the levels of negative reaction were not as
discrepantcomparing verbal to touching behavior when John and Jane had frequently socialized.
Finally, because 25 per cent of our sample (n = 263) were graduate or professional students, we
compared their responses to the responses of undergraduates.First, we examined main effects of
graduate student status and found that graduate students perceived higher levels of controlling
behavior and had more negative reactions and fewer positive reactions, compared to
undergraduates. The effect sizes explained less than 2 per cent of the variance. Next, we
examined interactions of graduate student status with the independent variables. These analyses
revealed no interactions, indicating that the main effects in our study did not vary for graduate
versus undergraduate students.
To summarize the results, we found that: (1) when subjects evaluated scenarios in which the
male and female had frequently socialized, there was less of a perception of controlling/power
behavior, fewer negative reactions, less interpretation of the male's behavior as negative, and
more positive reactions; (2) in evaluating the scenarios, subjects saw sexual behavior from a peer
as less controlling and power-oriented compared to behavior from a supervisor or executive. No
other differencesfor status emerged;(3) verbalcommentsresulted in subjects' perceptions of more
controlling/power behavior, fewer positive reactions, and a more negative interpretation of the
male's behavior, compared to touching behavior; (4) sexual behavior in the primary worksetting
was perceived as more controlling/power oriented but equally offensive in other types of
reactions, compared to behavior in the bar setting; (5) there were no differencesin perceptions for
male compared to female subjects, neither were there any interactions of sex with other factors;
(6) subjects' previous experience with sexual harassment was related (weakly) only to perceptions
of controlling/power behavior; (7) most interactions among independent variables were not
significant, the exception being touch/verbal with amount of previous socializing.
WORKPLACE SEXUAL HARASSMENT 499

Discussion

Sexual harassment is a serious problem with important ramifications, both for individuals and for
organizations (York and Brookhouse, 1988). This study provides additional clarity concerning
contextual factors which contribute to employees' evaluations of behavior which is ambiguous,
but which may be seen as sexual harassment. Our research is embedded in the overall context of
role theory which suggests that people are expected to behave according to the roles they fulfill
(Katz and Kahn, 1978), and that behavior from others is interpreted in light of these role
expectations. Our findings, based upon the perceptions of a large sample of respondents, can
contribute to the further understanding of models of sexual behavior and harassment at work. Of
the six hypotheses, four received partial or strong support, and two were not supported. The
results indicate that, contrary to our prediction, verbal behavior was perceived as more
controlling and was less likely to be interpreted positively. This finding is contrary to previous
research, such as the study by Gutek and colleagues (1983), in which touching was viewed more
negatively than verbal behavior. However, there can be many varieties of touching. In the Gutek
study touching was a 'pat on the fanny', which may be more explicitly sexual than in our study,
which involved a male putting his arm around a female's shoulders. An arm around the
shoulders may be a less sexual and less offensive touching behavior, but probably also a more
common behavior at work. Additional studies can identify more precisely which types of
touching are the more offensive at work.
The second finding directly contradicting our prediction was the lack of any significant male-
female differencesamong respondents in evaluation of the potentially harassing behavior. This is
perhaps our most striking finding, given that most previous research has found males to be less
offended and more accepting of sexual behavior (Fitzgerald and Shullman, 1993), especially in
circumstances which are the most ambiguous (Ken and Ryan, 1986). The lack of differences in
our study suggests some interesting trends. In the decade of the 1980s, increasing attention was
paid to sexual harassment in all aspects of our society, especially at work. In addition to a
number of celebrated court cases, many organizations introduced sexual harassment training
programs, often designed with males as the target audience. This increasing awareness may have
had an especially strong effect on male sensitivity to harassment issues, and to males changing the
way they define sexual harassment. This heightened sensitivity may be more pronounced in a
university community and among students, such as the respondents in our study. Given the
lower response rate from men compared to women, it is also possible that men who considered
sexual harassment as unimportant were less likely to respond to the survey.
Male-female perceptual differences may also be related to sex-type of respondents, such as
androgynous, same-sexed, or cross-sexed, which was not included in our study (Sheffey and
Tindale, 1992). We await future studies which may confirm a diminishing of this particular type
of gender gap between males and females in the workforce. Research evaluation of sexual
harassment programs and interventions could also be extremely valuable.
Our results also clarify employees' interpretation of behavior from a higher (job) status
person. In our study behavior from a peer (equal job status) was perceived as less controlling and
power-oriented than from a direct supervisor or executive. Future sexual harassment research
would benefit from a more comprehensive examination of the role of power, broadly construed,
in sexually harassing behavior and responses (Tinsley and Stockdale, 1993). As discussed earlier,
researchers should also study social status and harassment; as Littler-Bishop et al. (1982)
suggested, employees may be more tolerant of harassing behavior from those with higher social
status.
500 T. W. DOUGHERTY ET AL.

We found little relationship of previous experience of sexual harassment at the University with
our dependent variables. We acknowledge, however, that we did not measure harassment
experience outside the University, which could have had adaptive or sensitizing effects. We also
acknowledge that our measure of past experience did not clearly distinguish sexual behavior
resulting in negative consequences for the recipient. We should also point out that we used a
quite comprehensive set of dependent measures reflecting perceptions of potentially harassing
behaviors. A variety of dependent variables gives a fuller picture of the effects of treatments,
especially since treatments sometimes affect dependent variables quite differently. We also used
response items including terms such as 'offensive work environment', 'hostile or intimidating
work environment' and 'interferedwith work performance' (see Table 1), which are key terms in
the EEOC definition of sexual harassment. We did not, however, directly ask respondents if
sexual harassment occurred. Even though our focus was not on legal or personal definitions of
harassment, this omission could be considered a limitation of our study. Related to this issue,
Fitzgerald and Shullman (1993) note that there are many definitional and measurement
problems in conceptualizing and measuring sexual harassment. They suggest, for example, that
the words 'sexual harassment' be avoided when measuring the incidence of actual experienced
behaviors.
Our exploratory analyses of interactions among independent variables (including respondent
sex) produced few significant findings. The interaction of touching with previous socializing
suggests that the nature of the relationship between a man and a woman is important in
determining how work behavior is interpreted. This issue has practical implications (e.g. for
harassment training programs) and deserves more research attention.
This study analyzed perceptions of university students, which can be considered as a limitation
in generalizing the results to the work setting. These students, however, are likely to have seen
and dealt with social-sexual behavior many times in their past work and school environments.
The respondent group also included a sizable number of graduate and professional students,
many of whom had extensive previous work and professional experience. Researchers who have
the opportunity to study social-sexual behavior in the context of actual work settings will provide
valuable contributions.

References

Brewer, M. B. (1982). 'Further beyond nine to five: An integration and future direction', Journal of Social
Issues, 38(4), 149-157.
Bursik, K. (1992). 'Perceptions of sexual harassment in an academic context', Sex Roles, 27, 401-412.
Cleveland, J. N. and Kerst, M. E. (1993). 'Sexual harassment and perceptions of power: An under-
articulated relationship', Journal of Vocational Behavior, 42, 49-67.
Cohen, A. G. and Gutek, B. A. (1985). 'Dimensions of perceptions of social-sexual behavior in a work
setting', Sex Roles, 13, 317-327.
Collins, E. G. C. and Blodgett, T. B. (1981). 'Sexual harassment ... some see it ... some won't', Harvard
Business Review, 59(2), 76-95.
Fitzgerald, L. F., Hulin, C. L. and Drasgow, F. (1992). 'Predicting outcomes of sexual harassment: An
integrated process model'. Paper presented at the 2nd APA/NIOSH Conference on Stress and the
Workplace, Washington, D.C., November.
Fitzgerald, L. F. and Ormerod, A. J. (1991). 'Perceptions of sexual harassment: The influence of gender
and academic context', Psychology of Women Quarterly, 15, 281-294.
Fitzgerald, L. F. and Shullman, S. L. (1993). 'Sexual harassment: A research analysis and agenda for the
1990s', Journal of Vocational Behavior, 42, 8-27.
WORKPLACESEXUALHARASSMENT 501

Gutek, B. A. (1985). Sex and the Workplace,Jossey-Bass, San Francisco.


Gutek, B. A., Morasch, B. and Cohen, A. G. (1983). 'Interpretingsocial-sexual behavior in a work setting',
Journal of Vocational Behavior, 32, 30-48.
Jackson, S. E. and Schuler, R. S. (1995). 'Understanding human resource management in the context of
organizations and their environments', Annual Review of Psychology, 46, 237-264.
Katz, D. and Kahn, R. L. (1978). The Social Psychology of Organizations,2nd edn, Wiley, New York.
Kenig, S. and Ryan, J. (1986). 'Sex differences in levels of tolerance and attribution of blame for sexual
harassment on a university campus', Sex Roles, 15, 535-549.
Konrad, A. M. and Gutek, B. A. (1986). 'Impact of work experience on attitudes toward sexual harass-
ment', AdministrativeScience Quarterly, 31, 422-438.
Littler-Bishop, S., Seidler-Feller, D. and Opaluch, R. E. (1982). 'Sexual harassment in the workplace as a
function of initiator's status: The case of airline personnel', Journal of Social Issues, 38(4), 137-148.
Marks, M. A, and Nelson, E. S. (1993). 'Sexual harassment on campus: Effects of professor gender on
perception of sexually harassing behaviors', Sex Roles, 28, 207-217.
Mazer, D. B. and Percival, E. F. (1989). 'Idealogy or experience? The relationships among perceptions,
attitudes, and experiences of sexual harassment in university students', Sex Roles, 20, 135-147.
Reilly, T., Carpenter, S., Dull, V. and Bartlett, K. (1982). 'The factorial survey: An approach to defining
sexual harassment on campus', Journal of Social Issues, 38(4), 99-1 10.
Sheffey, S. and Tindale, R. S. (1992). 'Perceptions of sexual harassment in the workplace', Journal of
Applied Social Psychology, 22, 1502-1520.
Tangri, S. S., Burt, M. E. and Johnson, L. B. (1982). 'Sexual harassment at work: Three explanatory
models', Journal of Social Issues, 38(4), 33-54.
Tinsley, H. E. A. and Stockdale, M. S. (1993). 'Sexual harassment in the workplace', Journal of Vocational
Behavior, 42, 1-4.
U.S. Merit Systems Protection Board (1981). Sexual Harassment in the Federal Workplace.Is it a Problem?
U.S. Government Printing Office, Washington, D.C.
Weber-Burdin, E. and Rossi, P. H. (1982). 'Defining sexual harassment on campus: A replication and
extension', Journal of Social Issues, 38(4), 111-120.
York, K. M. and Brookhouse, K. J. (1988). 'The legal history of work-related sexual harassment and
implications for employers', Employee Responsibilitiesand Rights Journal, 1, 227-237.

Você também pode gostar