This study examined attributions of responsibility in cases of sexual harassment. It found that women who endorsed traditional gender roles were more likely to blame victims of sexual harassment. The study also found that male observers assigned more responsibility to female victims than female observers or victims did. Additionally, women who had experienced sexual harassment themselves were less likely to blame victims than women who had not been victims. The research suggests that victims sometimes engage in self-blame after experiencing sexual harassment.
50 ĐỀ LUYỆN THI IOE LỚP 9 - NĂM HỌC 2022-2023 (CÓ LINK HÌNH, FILE AUDIO VÀ ĐÁ...
Attributions And Assignment Of Responsibility In Sexual Harassment
1. Journulof zyxwvutsr
Sociallssues, Vol. zyxwvuts
38, No. 4, 1982, zyxwvu
pp. 121-136 zyxwv
Attributions and Assignment of Responsibility
in Sexual Harassment
Inger W. Jensen
University zyxwvutsr
of Culifornia uf Los Angeles zyxwvu
Barbara A. Gutek
Clarernonr Gruduute School
Although research has investigated the extent to which sexual harassment
exists and what people are or are no1 doing about it, little has been written
about the personal attributions made for these behaviors by either those
who have or those who have not been sexually harassed. Research on
rape and the assignment of responsiblity for a rape may be useful as a base
from which to generalize about sexual harassment and the various ways in
which victims and viewers respond to it. This study consists of data from
an initial survey investigating the occurrence of sexual harassment at the
workplace and follow-up interviews with female workers who reported in
the initial survey having been sexually harassed. The research focused on
the application of attribution theory in four areas: 1) sex differences in
assignment of responsibilityfor sexual harassment, 2) effects of having ex-
perienced sexual harassment on assignment of responsibility, 3) self-blame
in cases of sexual harassment and its affectiveand work-related effects, 4)
the role of sex-role beliefs on assignment of responsibility to self and others.
One of the stronger findings of the study concerned the role of sex-role
beliefs. Women who have traditional sex-role beliefs are more likely to
blame themselves and other victimsfor being sexually harassed. zyx
The research reported here was partially supported by a National lnrtitute of Mental
Health (NIMH) grant UPHS-MH-32606-01 lo the second author. We would like to thank
Marilynn Brewer, Joe McGrath, and an anonymous reviewer for their comments and sug-
gestions. We would also like lo acknowledge the help of Tom Wickens in providing advice
on the log linear analyses.
Correspondence regarding this article should be addressed to 1. W. Jensen, Cedars Sinai
Medical Office Towers, 8635 W. Third zyxwvu
St., Suite 770W, Los Angeles, California 90048.
121
0022~S37/82/1200-0121$03.00/1(s) 1982 Society for the Psychological Study of Social Issuer
2. 122 zyxwvutsrq
lnger W. Jensen and Barbara A. Gulek zy
Although research has investigated the extent to which sexual harass-
ment exists and what people are or are not doing about it, little has been
written about the personal attributions made for these behaviors by either
those who have or those who have not experienced incidents of sexual
harassment.
Benson and Thomson (1982) have suggested that a significant gap
appears to exist between men and women in their perception of the fre-
quency and type of harassment that actually occurs at the workplace.
They also reported considerable sex differences in the perception of the
difficulty a woman would have in handling an unwanted sexual advance.
Vogelmann-Sine, Ervin, Christensen, Warmsun, and Ulmann (1979) noted
the importance of knowing how behaviors are interpreted in complex
interpersonal situations in order to avoid possible misunderstanding of
intent.
The literature on sexual harassment indicates that sex differences
exist in the perception of what constitutes harassing behaviors (Gutek,
Nakamura, Gahart, Handschumacher Jensen, zyxw
& Russell, 1980; Gutek,
Note 1). Although the theoretical bases for these different perceptions of
sexual harassment have not yet been addressed, the rape literature suggests
a variety of theories for such sex differences in attributions of causality
and assignment of responsiblity.
Although sexual harassment and rape are certainly not the same, they
do have some common characteristics. In each case, one person is trying
to force his sexual attentions on an unwilling other. While in rape the
coercion usually is the threat of bodily harm, in sexual harassment it
usually is the threat of depriving individuals of their jobs or job op-
portunities.
Groth (1979) suggested that there are three basic ways in which a
person gains sexual access to another person: 1) through consent (negotia-
tion); 2) through pressure (exploitation); or 3) through force (intimidation).
He further noted that the difference between pressure and force is that
in the pressured situation, the victim is sexually harassed or exploited,
while in the forced assault, the victim is raped. He suggested that there
is enough similarity in the underlying factors of the two situations to discuss
both meaningfully under the single term of rape. In further support of this
notion, Medea and Thompson (1974) asserted that rape could be construed
as a continuum of various degrees of violence with varying degrees of
sexual intimacy.
Although the various theories about attributional processes make
some conflicting predictions, they do have a number of things in com-
3. Attributions zyxwvutsrq
123 zy
mon. Most predict that sex differences will exist in attributions of locus
of responsibility as well as in how the incidents will be viewed. The study
described in this article addressed these issues. It investigates how well
various attribution theories account for the assignment of responsibility
for sexual harassment incidents, and it explores the impact of sex-role
beliefs on such assignment of responsibility.
DESIGN OF THE STUDY
The data addressing sex differences in assignment of responsibility
were derived from interviews with a representative sample of workers in
Los Angeles County with women oversampled; 827 women and 405 men
were interviewed (see Gutek, Note 1, for a description of this survey).
All other analyses use data obtained from 135 of the female respondents
who were re-contacted and who will hereafter be referred to as the victims
of sexual harassment.
The pool of respondents for the follow-up interviews consisted of
293 women who, in the initial survey, agreed to be recontacted and who
had reported experiences of one or more of the following sexually harassing
behaviors at their present or previous jobs:
a. having been touched by a man in a sexual way on the job;
b. being expected to go out with a man with the understanding that it
would hurt her job situation if she refused, or would help if she accepted;
c. being expected to engage in sexual relations with a man with the
understanding that it would hurt her job situation if she refused, or
help if she accepted.
The telephone interviews in this follow-up study were conducted
with the help of a computerized telephone system that enabled direct
programming of responses (Shure zyxwv
& Meeker, 1978). Due to the lapse in time
since the original contact (approximately 15 months), a substantial number
of potential respondents could not be contacted; 135 interviews were com-
pleted, close to a completion rate of 50%.
The interviews were conducted by college seniors who had received
prior training. In general, the interviews lasted between 15 and 20 minutes.
Once the subject’s consent to be reinterviewed was obtained, the interviewer
described the incident cited previously by the subject to see that the subject
still remembered the incident. All subjects were asked the same questions.
The first seven items dealt with feelings the respondent may have had
4. 124 zyxwvutsrq
lnger W. Jensen and Barbara A. Gutek zy
following the incident. Then, some questions designed to ascertain locus of
responsibility were presented, followed by a number of questions about
reporting the incident and possible reasons for not reporting it. This was
followed by questions about attitudes towards sexual harassment. Finally,
a shortened version (17 items) of a scale measuring feminist attitudes
(Smith, Ferree, zyxwvu
& Miller, 1975) was presented. The scale contained items
assessing beliefs about women’s personality traits and what constitutes
appropriate role behavior in women.
The main classes of variables from the follow-up study that are
relevant to this paper are: 1) attribution items measuring self-blame and
other-woman blame, 2) behavior related to reporting the incident, 3)
affect variables (reports of feelings of anger, sadness, hurt, anxiety,
disgust), 4) job-related outcomes, 5) health-related outcomes, 6) sex-role
beliefs (Smith et al., 1975).
RESULTS
The results are divided into four sections: sex differences in assign-
ment of responsibility to female victims of sexual harassment; the influence
of being a victim on the assignment of responsibility; an exploration of
the phenomenon of self-blame among victims; the relation between sex-
role beliefs and assignment of responsibility.
Sex Differences in Assignment of Responsibility
Several authors (Shaver, 1970; Walster, 1966) have suggested that
the differential motivations of an actor and an observer may be an im-
portant factor in assignment of responsibility to a victim. Walster (1966)
found that as the consequences of an accident grew more severe, there
was a tendency for observers to assign increasing responsibility for the
event to the victim rather than to chance. She hypothesized that in the
case of a minor accident, the observer may feel sympathetic toward the
victim and attribute the responsibility for the outcome to either chance or
other persons. However, in the case of a severe catastrophe, ascribing its
occurrence to chance would leave observers with the uncomfortable thought
that a similar thing might happen to them. By stressing the dissimilarity
of the victim to themselves, observers may disassociate themselves from
a similar tragedy.
Shaver (1970) introduced the notion that such a desire to avoid fear
of unfortunate outcomes may be altered by the need for observers to
5. Attributions zyxwvuts
125 zy
avoid blame for such an incident and consequently may lead to differential
assignment of causal attributions and responsibility depending on the ob-
server’s perceived likelihood of being involved in a similar incident. He
further made a distinction between relevance due to personal similarity
between the observer and the victim of the accident, and relevance due
to situational similarities between them. He suggested that situational
similarity was a necessary factor, and once it was present, the degree of
personal identification would determine which attributional system would
take place. In “harm avoidance,” indicative of low personal relevance,
increased personal responsibility is assigned to the victim; in “blame
avoidance,” indicative of high personal relevance, decreased responsibility
is assigned in order for the observer to avoid being blamed should a similar
unfortunateincident happen to hcr.
A different motivational theory was suggested by Lerner (1965, 1971,
1978). He postulated that people were motivated to believe in a “just
world” where people get what they deserve, good or bad. To believe other-
wise would raise the question of whether the person can trust his or her
own environment. This “just world’’ notion suggests that when people
hear of an unfortunate incident, they would want to believe that the
victim deserved whatever suffering he or she experienced.
When applied to incidents of sexual harassment, the theories make
differential predictions about the assignment of responsibility by male and
female observers. According to Shaver’s defensive notion, male observers
should assign more responsibility for such incidents to female victims than
would female observers. Walster’s defensive notion predicts that female
respondents should view the female victim of sexual harassment as more
responsible than should the male respondents. In contrast to both of those
theories, Lerner’s “just world” would suggest no differences between male
and female respondents in the assignment of responsibility.
A review of the research investigating attribution theory and rape
suggests that Shaver’s defensive notion best accounts for sex differences in
assignment of responsibility for such incidents (cf. Calhoun, Selby, Cann, zy
& Keller, 1978; Selby, Calhoun, & Brock, 1977). It is reasonable to suggest
that male respondents, not wanting to be blamed in the future, would
assign more responsibility to the women, and similarly, the women, not
wanting to be blamed should such an event happen to them in the future,
would assign less responsibility to the women and more to the men. Shaver’s
version of attribution theory also best explains the data about the sex
differences in assignment of responsibility in our initial survey on sexual
harassment. More responsibility was assigned to the victim of sexual
harassment by the male than the female respondents in the large survey.
Significantly more of the male than female respondents endorsed the state-
ment, “When a woman is asked by a man at work to engage in sexual
6. 126 Inyer zyxwv
W. Jensen and Barbara A. Gutek zy
relations, it’s usually because she did something to bring it about” (54.5%
of males versus 44.7% of females; zyxw
x’ zyxwv
= 9.41, p < .001).
The Influence of Being a Victim on the Assignment of
Responsibility
Attribution theory also addresses the differential assignment of
responsibility as a function of having been a victim. When applied to
incidents of sexual harassment, attribution theory suggests that women in
general will attribute less blame than men to a victim of sexual harassment.
However, significant differences in assignment of responsibility should
still emerge in a comparison between those women who had and those
who had not themselves been sexually harassed.
This notion was supported by the responses of the women inter-
viewed in the larger study. Although overall the majority of the female
victims agreed with the statement, “Women who are asked by men at work
to engage in sexual relations could have done something to prevent it”
(62% agreed), the agreement was significantly greater among those women
who had zyxwvuts
not been sexually harassed (74% agreed) (x’ = 10.8, df = 1,
p < .01). A difference in the same direction, although not statistically
significant, was found in respondents’ endorsement of the statement,
“When a woman is asked by a man at work to engage in sexual relations,
it’s usually because she did something to bring it about”; 46.4% of non-
victims and 42.2% of victims agreed.
The data further show that if the victims believed that a woman
does something to bring about an incident of sexual harassment, they
were also highly likely to believe that the woman could do something to
prevent such incidents. The relationship between the two variables (women
could prevent and women bring about) was significant (x’ = 290, df =
9, p < .001). It was, however, asymmetric. The female victims who strongly
agreed with the statement, “Women do something to bring it about,”
were highly likely to also agree with the statement, “Women could do
something to prevent it” (92%). The same strong relationship was not
found the other way around (59%). That a woman is expected to prevent
things from “going too far” is consistent with traditional sex-role beliefs,
which will be discussed later in this paper.
Self-Blame
The attributions made by victims about their own situation has been
a focus of interest in the rape research and will be the focm of the remainder
7. Attributions 127 zy
of this paper. Attribution theory applied to rape situations suggests that
victims experience self-blame (e.g., Bryant zyxw
& Cirel, 1977; Burges &
Holshom, 1974). Janoff-Bulman (1979) suggested that two different kinds
of self-blame exist: control related self-blame and esteem related self-
blame. With the former self-attributional strategy, persons blame the
unfortunate incident on their own behavior, such as having engaged in
a certain activity, or inversely, having failed to do so (e.g., I should not
have gone down this street alone at night; I should have kept my door
bolted). In the latter type of self-blame, the persons fault their own
character (i.e., I am just a careless person). While the self-blame focused
on behavior is termed “behavioral,” the esteem related self-blame is
termed “characterological” by Janoff-Bulman.
While there has been a scarcity of research dealing with this issue,
it seemed plausible that there may also be such assignment of respon-
sibility to self by victims of sexual harassment. Our follow-up study ad-
dressed several questions regarding self-blame: 1) will a victim of sexual
harassment engage in self-blame? 2) will a victim who engages in self-blame
be less likely than other victims to seek help? 3) what are the affective
consequences of self-blame and what further effects does self-blame have?
Three statements designed to measure the presence of self-blame
were presented to the women in the follow-up survey: zyx
1) “Perhaps
something in my behavior brought it about,” 2) “Perhaps something in
my behavior may have encouraged the man,’’ and 3) “I am the sort of
person these kinds of incidents are likely to happen to.” The first two
statements differ somewhat from the latter, in that they dealt with the
respondents’ behavior; the correlation between the two is .53. The “sort
of person” statement, on the other hand, ascribed the incident to more
characterological qualities of the respondent and yielded low correlations
with the two behavioral attributions (.17and .12 respectively). The majority
of victims did not evidence self-blame; only 25.4% of the 135 victims
agreed with the first statement, 29.3% with the second statement, and
20.9% with the third. It is possible, of course, that greater self-blame
may be evidenced when assessed soon after an incident; all of these in-
cidents occurred at least 15months before the interview. zyx
Self-Blame and Incident Reporting
It was hypothesized that a victim displaying self-blame would be
less likely to have reported the incident to someone in authority. To in-
vestigate the relationship between self-blame and help-seeking behavior,
responses to the following questions were examined: 1) “Did you report
8. 128 zyxwvutsrq
lnger W. Jensen and Barbara A. Culek zy
your experience to anyone in authority like a supervisor, personnel director,
attorney, etc.?” and 2) “Did you talk to anyone you were close to about
the experience, like family, a friend, or a co-worker?” As predicted, those
who endorsed the two behavioral self-blame items were less likely to
have reported the incident (Table 1). The Pearson correlation coefficients
show significant relationships between the two behavioral self-blame items
and reporting the incident, but failed to support such a relationship for
the characterological self-blame item. The same relation was found between
the self-blame items and the item investigating whether victims had talked
to friends about the incident. Table 1 indicates that respondents who had
endorsed the behavioral self-blame statements tended not to have talked
to anyone about the incident of sexual harassment.
One possible reason for the lack of a relationship between the
characterological self-blame item and reporting the incident may be that
this item does not adequately measure self-blame but rather may reflect
feelings of being a victim. Another factor is that the characterological
self-blame item is psychometrically weak because the response distribu-
tion is so skewed.
One of the ways of explaining the non-report of incidents among the
women who engage in behavioral self-blame is the process of victimiza-
tion of women suggested by Weis and Borges (1973) and reported ex-
tensively in the rape literature. This notion proposes that other people
will hold a victim responsible for her rape or sexual harassment. As part of
this process, the victim internalizes the blame and will try to find some-
thing in her behavior to which she can attribute the incident. It is then
not surprising that she will hesitate to report an incident to authorities or
even to friends. As further support of this notion, more than one-third
of the victims endorsed the following statement as a reason for non-report
of the incident: “I thought it would be held against me or that I would
be blamed.”
Table 1. The Relationship Between Self-Blame and Help-Seeking Behavioi
Agreement with statements Reported incident Talked about incident
Behavioral self-blame:
behavior brought it
about. zyxwvutsr
. .
may have encouraged the man. . .
Perhaps something in my -0.22** -0.25**
Somethingabout my behavior -0.19** -0.30***
Characterological self-blame:
I am the sort of person these
kinds zyxwvutsr
of incidents are likely to
happen to . . .
0.04 --0.10 zyx
*p < zyxwvuts
O.O5;**p < 0.01; ***p < 0.001.
9. Allribulions zyxwvutsr
129 zy
The data in the follow-up survey also show that if a victim blames
herself, she is concerned about protecting the harasser. Positive correla-
tions were obtained for the relationships between all three self-blame items
and endorsement of the statement, “I did not want to hurt the person
who bothered me.” (Correlations for “behavior brought it about,” “be-
havior encouraged,” and “sort of person” were .19, .33, and .23 respec-
tively.) zyxwvuts
SelfBlame zyxwvut
and Affect
Janoff-Bulman (1979) suggested the efficacy of distinguishing be-
tween characterological and behavioral self-blame because she felt they had
different consequences. She contended that characterological self-blame
leads to feelings of worthlessness, of deserving one’s fate, and especially,
to depression. The hypothesized link between characterological self-
blame and reports of depression was not upheld in this study. Neither of
the two types of self-blame were related to reports zyx
of depression. The
coefficients between the self-blame items and depression were all under .lo.
A relatively small percentage of respondents (20%) gave reports of
depression in response to the sexual harassment incident, in contrast to
reports of other types of affect such as anger (68%) and disgust (80%).
Possible explanations may be that feelings of depression are more difficult
to relate to a stranger and/or harder to remember after more than a year
has passed. It is possible that the college students who evidenced depres-
sion in Janoff-Bulman’s study might not have revealed having experienced
such affect in a telephone interview over one year later.
In this study, seven different affective responses of victims to their
sexual harassment were assessed. A factor analysis of the items yielded
three factors. One included the affective responses of hurt, sadness, and
depression, and was labeled inward-directed affect; the second, labeled
outward-directed affect, was formed by the two responses, anger and
disgust. The third factor consisted of the single item, anxiety. Reliability
analyses supported combining the inward-directed affect items (Cronbach’s
alpha zyxwvut
= .72) as well as the two outward-directed affect items, anger
and disgust (Cronbach’s alpha = .80).
Janoff-Bulman’s (1979) research would suggest that a positive rela-
tionship might exist between behavioral self-blame and outward-directed
affect and also between characterological self-blame and inward-directed
affect as well as anxiety. In order to test this, the two items measuring
behavioral self-blame were combined into a single index (Cronbach’s
alpha = .64). There were significant relationships between behavioral
self-blame and outward-directed affect, and also between characterological
10. 130 zyxwvutsr
lnger zyxwv
W. Jensen and Barbara A. Gutek
Table 2. zyxwvuts
The Relationship Between Type of Blame for Incidents
of Sexual Harassment and Type of Affect Experienced
Pearson correlation coefficients
Type of affect
T w e of blame Inward Outward Anxiety zyx
~~ ~
Behavioral self-blame 0.02 0.27*** 0.04
Characterological self-blame 0.01 0.03 .15** zyxw
*p < 0.05; zyxwvuts
**p < 0.01; ***p < 0.001.
self-blame and anxiety. However, the expected relationship between
characterological self-blame and inward-directed affect was not obtained
(Table 2). This may be because the characterological self-blame item
is not normally distributed and apparently has low reliability.
Effects zyxwvu
of Negative Affect Resultingfrom Sexual Harassment
In the case of sexual harassment, the knowledge that victims have
negative emotional reactions is not as likely to impress employers as the
fact that those emotional reactions have work-related consequences. Several
questions were asked in the follow-up interview to ascertain whether
feelings towards work had been adversely affected. Analyses revealed
significant relationships between reports of feelings of the negative affect
items-hurt, sadness, depression, anxiety, anger, and disgust -and an item
measuring negative attitude toward the job -loss of motivation, feeling
distracted, and dreading to go to work.
If a victim reported having experienced any of these types of negative
affect, she was also likely to report having experienced loss of motivation
and/or feeling distracted at the workplace zyxw
(F = .27). There also was a
significant relationship between negative affect and an item measuring
respondents’ physical well-being after the incident. If she had experienced
any of the types of negative affect, she was also likely to have experienced
negative physical effects (headaches, nervousness and sleeplessness) (f =
.26).
Sex-Role Beliefs and Assignment of Responsibility
Several studies on rape (Brownmiller, 1975; Field, 1978; Medea &
Thompson, 1974; Smith, Keating, Hester, & Mitchell, 1976; Krulewitz &
Payne, Note 2) conclude that general attitudes toward women, as in-
corporated in traditional versus pro-feminist views, are the best predictors
11. Atlributions 131 zy
of attitudes about rape. Krulewitz and Payne (Note zyx
2) found support
for the notion that subjects’ attitudes toivard traditional sex-roles may
influence their perceptions about rape. While normative behavior by
women is often described as “submissive, passive, not aggressive, and
fearful,” male behavior is frequently described as “dominant, aggressive,
adventurous, opportunistic, and proud of sexual ability” (Heilbrun, 1976).
Krulewitz and Payne (Note 2) suggested that the rapist could be viewed
as displaying an extreme version of such stereotypically masculine sex-
role behavior; i.e., sexually active, aggressive, and opportunistic. They
suggested that this may lead to perceiving the rape as normative male-
female sexual behavior, especially when force and physical violence is
low. Thus, an incident will be characterized as rape only when the per-
petrator uses substantial physical force. They proposed, however, that
this would only hold true for subjects incorporating more traditional
sex-role beliefs; those with nonstereotypical attitudes towards women
would be likely to view an incident as rape irrespective of the amount of
force used or the subject’s physical condition after the rape. Labeling
those holding the more traditional views towards women non-feminist
and those with stronger positive attitudes towards feminism pro-feminist,
Krulewitz and Payne found that while the amount of force depicted did
not affect the attributions of certainty by the pro-feminist female subjects,
non-feminist female subjects required substantial evidence of force before
they were likely to judge an incident as a rape. As the level of force in-
creased, victims were viewed as more helpless, but also, as more respect-
able. This indication that a respectable woman is expected to withstand
rape unless violent force is exerted by the rapist has been supported by
other studies (Brownmiller, 1975; Medea zyxw
& Thompson, 1974). They argue
that commonly held societal attitudes dictate that women should be
responsible for protecting their sexual purity. In further support, Field
(1978) found that pro-feminist subjects assigned less responsibility to the
victim than did subjects who embraced more traditional, or conservative
sex-role beliefs.
The possible impact of sex-role beliefs on perceptions of sexual
harassment has not been addressed. Earlier the victimization process was
suggested as one possible explanation for self-blame. Another exp1anatio:i
is that sex-role beliefs affect self-blame. In order to investigate whether
the effects of sex-role beliefs could be generalized from research on rape
to sexual harassment, the 135 women in the follow-up survey who had
reported their experiences of sexual harassment were administered a sex-
role scale (Smith et al., 1975) along with the other questions. This scale
used 5-point Likert-type items to measure acceptance or rejection of central
beliefs of feminism and the acceptance or rejection of traditional sex-
role beliefs.
12. 132 lnger zyxwv
W. Jensen and Barbara A. Gutek zy
The literature on rape suggests that more assignment of respon-
sibility to victims will be made by those who adhere to traditional sex-
role beliefs. Generalizing this to incidents of sexual harassment suggested
that of those women who reported experiences of sexual harassment, those
who adhere to traditional stereotypic sex-role concepts would assign greater
responsibility to women in general for incidents of sexual harassment.
Three questions dealing with assignment of responsibility to women
in general for sexual harassment were fermed into a single index (Cronbach’s
alpha zyxwvut
= .60). The three items concerned whether women provoke ad-
vances, do something to bring them about, and could do something to
prevent them. There was a significant correlation between the responses to
this index and to the sex-role scale zyxwv
(r = .37), indicating that victims who
obtain traditional scores on the sex-role scale are more likely than those
with non-traditional scores to hold women responsible for incidents of
sexual harassment.
Sex-Role Beliefs and Self-Blame
Sex-role beliefs should also have an impact on whether or not a
victim of sexual harassment will blame herself. In support of this con-
tention, there were significant correlations between sex-role beliefs and the
index containing the two behavioral self-blame items (r = .26) and be-
tween sex-role beliefs and characterological self-blame (r = .33).
Relation Between Self-Blame and Blaming Others
Sex-role beliefs of victims are related to both self-blame and blaming
other women, suggesting that there may be a relationship between self-
blame and other-woman blame. We predicted a stronger tendency to blame
others for the women who engaged in self-blame (behavioral and character-
ological) in contrast to those who did not evidence self-blame. Con-
sistent with this prediction were significant positive correlations between
the index assessing blame to other women and both characterological
self-blame (r = .23) and the two-item index measuring behavioral self-
blame (r = .32).
Thus, sex-role beliefs, blaming others, and self-blame are all related.
What are the causal or sequential processes by which they are linked?
One possibility is that both other-woman-blame and self-blame are con-
sequences of sex-role beliefs. This possibility can be directly tested by
examining them in a three-dimensional test of frequencies, that gives the
difference in the fit between two models (cf. Fienberg, 1980). One model
contains only direct relationships between self-blame and sex-role beliefs
13. Allribuliuns 133 zy
sex-role belief
Fig. zyxwvuts
1. Models of relationships among sex-role beliefs, self-blame, and other-
woman-blame.
and between sex-role beliefs and other-woman-blame (Figure zyx
1, excluding
dashed line); the other model contains all three relationships -sex-role
beliefs and other-woman-blame, sex-role beliefs and self-blame, and
self-blame and other-woman-blame (Figure 1 with the dashed line).
The results of the model testing indicate that the relationship between
the two forms of blame can be attributed to their common association
with sex-role beliefs, and not to any direct causal links between them.'
These results are not consistent with the belief that victimization is a
consequence of internalizing others' perceptions. If that were the case,
there should have been an independent effect for the relation between
other-woman-blame and self-blame. The results of this analysis do suggest
a mechanism for reducing self-blame -acquiring more feminist sex-role
beliefs.
CONCLUSIONSAND IMPLICATIONS
The hypothesis that men and women differ in their assignment of
responsibility regarding sexual harassment at the workplace is strongly
supported by our survey data. Men are more likely than women to blame
women for being sexually harassed. Among women, victims are less
likely than others to blame women for being sexually harassed.
Our data from the follow-up survey supported the research on rape
showing that victims engage in self-blame. Although the follow-up data
were collected over a year after the incident, between 20% and 30% of
women endorsed each of three self-blame statements. These victims showed
less characterological than behavioral self-blame.
'Formally, the test is of the conditional independence of self-blame [S] and other-
woman-blame [ 0 ] in the presence of the sex-role variable [R] (see Fienberg, 1980, chapter 3).
The model [RS] [RO] tests the relationships of sex-role to each of the two blame variables.
A fit of the model [RS] [RO] to the data gave a likelihood chi-square of 7.54 on 8 degrees
of freedom; a fit of the model [RS] [RO] [SO] gave a likelihood chi-square of 4.18 on
4 degrees of freedom. The difference between these two statistics tests the strength of the
self-other association. The difference gave a non-significant chi-square statistic ofzyx
3.36 with zy
4degrees of freedom.
14. 134 zyxwvutsrq
lnger W. Jensen and Barbara A. Gulek zy
Victims who evidenced behavioral self-blame were less likely than
other victims to report the incident or talk to someone about it. They
were also more likely to feel anger and disgust as a result of the incident,
yet they (as well as women who experienced characterological self-blame)
were more concerned about protecting the harasser than other victims
who did not report self-blame.
Moreover, women who had experienced negative affect were more
likely to report having experienced adverse physical reactions, as well as
such adverse job reactions as distraction and loss of motivation at the
workplace. This indicates that the occurrence of sexual harassment may
well be of substantital economic importance to the employer. In this regard,
these findings parallel those of a recent large-scale government study
investigating incidents of sexual harassment within the federal government
(U.S. Merit Systems Protection Board, 1981).
The results of the follow-up study also suggest that in trying to ac-
count for attributions and assignment of responsibility for experiences like
sexual harassment and rape, a salient factor to consider may be the accep-
tance of traditional versus progressive sex-role beliefs. The results of this
study suggest that sex-role beliefs will dictate how a woman who has
been sexually harassed is going to feel about the incident. If she has in-
corporated the traditional sex-role beliefs, she will be highly likely to blame
other women, as well as herself, for incidents of sexual harassment. Such
an attribution process, in return, will affect whether she reports the incident
to someone in authority or talks to co-workers or friends about it.
Furthermore, the data presented here show that sex-role beliefs have
direct effects on both blaming self and blaming others. These data are
in contrast to a model that shows a direct relationship between the two
kinds of self-blame. One common model proposes that self-blame is an
internalization of holding other women responsible for rape or sexual
harassment (cf. Weis zyxwvu
& Borges, 1973). In contrast, our data suggest that a
woman who is a victim of sexual harassment might engage in self-blame
because of her traditional sex-role beliefs rather than because she blames
other victims for being sexually harassed.
One implication of the impact of sex-role beliefs on attitudes towards
sexual harassment as well as rape is that a different approach toward
change may be warranted. Theorists in the past have suggested that at-
titudes could be changed by providing ample information about what
really happens in such incidents, and by changing specific beliefs about the
act and its consequences. It is possible that such an approach is not the
method of choice for these conditions, and that only by changing people’s
general sex-role beliefs can one effect a change in the attitudes towards
victims of harassment. Such a theory of attitudes and attitude change has
an immediate application in the legal profession, where much of the theory
15. Attributions 135 zy
about sexual harassment and debate about how it should be treated has
taken place (cf. MacKinnon, 1979). Our data suggest that, to the lawyer
who represents a victim of an incident, the most salient information to
obtain about the judge or the jury may not be their personal experiences,
or even their sex, but rather whether they endorse traditional or progressive
sex-role beliefs.
In general, this research project has demonstrated that sexual harass-
ment is sufficiently similar to rape to borrow theories and hypotheses.
In this case, the research on attribution theory that has been applied to
situations of rape proved to be a fruitful starting point for generating
hypotheses about sexual harassment. This ability to borrow from the
literature on rape is especially helpful since research and theory on rape is
better developed than research and theory on sexual harassment. A
continued interchange of research results between the two areas will even-
tually reveal crucial differences between rape and sexual harassment, as well
as the similarities here considered. This will yield benefits to both areas of
inquiry.
REFERENCE NOTES zyxw
1. Gutek, B. A. zyxwvutsr
Experiences zyxwvuts
of sexual hurussment: Resulis Jrotn a representutive survey.
Paper presented at the annual meeting of the American Psychological Association, LOS
Angeles, August 1981.
2. Krulewitz, J. E., & Payne, E. J. Sex differences in attributions about rupe, ruprsts, and
rape viclims. Paper presented at the meeting of the American Psychological Associa-
tion, San Francisco, 1977.
REFERENCES
Benson, D. J., & Thomson, G. Sexual harassment on a university campus: The confluence of
authority relations, sexual interest, and gender stratification. Social Problems, 1982,
29(3),236-251.
Brownmiller, S. Against our will: Men, women and rape. New York: Simon and Schuster,
197s.
Bryant, G., & Cirel, P. A communily response to rape: An exemplary projeci. (Polk County
Rape/Sexual Assault Case Center). Washington, D.C.: National Institute of Law
Enforcement and Criminal Justice, 1977.
Burgess, A. W., & Holmstrom, L. L. Rape trauma syndrome. American Journal of Psy-
chiatry,' 1974, 131,981-986.
Calhoun, L. G., Selby, J. W., Cann, A., & Keller, G. T. The effects of victim physical
attractiveness and sex of respondents on social reactions to victims of rape. The Brilish
Journalof Socialand Clinical Psychology, 1978, zyxwv
17, 191-192.
Field, H. S. Attitudes toward rape: A comparative analysis of police, rapists, crisis counselors,
and citizens. Journal of Personalily and Social Psychology, 1978, 36, 156-179.
Fienberg, S. E. The analysis of cross-classified culegorical data (2nd Ed.). Cambridge, MA:
MIT Press, 1980.
Groth, N. A. Men who rape: The psychology of the offender. New York: Plenum
Publishing Co., 1979.
16. 136 zyxwvuts
lnger W. Jensen and Barbara A. Gutek zy
Gutek, B. A,, Nakamura, zyxwvuts
C . Y., Gahart, M., Handschumacher Jensen, I. W., zyx
& Russell, D.
Sexuality in the workplace. zyxwvut
Basic and Applied Social Psychology, 1980, 1(3), 255-265.
Heilbrun, A. B., Jr. Measurement of masculine and feminine sex role identities as inde-
pendent dimensions. Journal zyxwvu
of Consulting and Clinical Psycliology, 1976, 44(2),
183-190.
Janoff-Bulman, R. Characterological versus behavioral self-blame: Inquiries into depression
and rape. Journal zyxwvutsr
of Personality and Social Psychology, 1979, 37, 1798-1809.
Lerner, M. J. Evaluation of performance as a function of performer’s reward and attractive-
ness. Journal ofPersonality and Social Psychology, 1965, I , 355-360.
Lerner, M. J. Observer’s evaluation of a victim: Justice, guilt, and verdical perception.
Jourrial of Personality and Social Psychology, 1971,20, 127-135.
Lerner, M. J., & Miller, D. T. Just world research and the attribution process: Looking
back and ahead. Psychological Bulletin, 1978,85, 1030-1051.
MacKinnon, C. Sexual harassment of working wornen. New Haven: Yale University Press,
1979.
Medea, A., & Thompson, K. Against rape. New York: Farrar, Strauss, and Giroux, 1974.
Selby, J. N., Calhoun, L. G. & Brock, T. A. Sex differences in the social perceptions of a
rape victim. Personalityand Social Psychology Bulletin, 1977,3, 412-415.
Shaver, K. G. Defensive attribution: Effects of severity and relevance on the responsibility
assigned to accidents. Journal ofPersonality and Social Psychology, 1970, 14, 101-113.
Shure, G. H., & Meeker, R. J. A minicomputer system for multiperson computer-
assisted telephone interviewing. Behavior Research Methods and lnstruriientation,
Smith, E. R., Ferree, M. M., & Miller, F. D. A short scale of attitudes towards feminism.
RepresentativeResearcli in Social Psychology, 1975, 6, 5 1-56.
Smith, R. E., Keating, J. P., Hester, R. K., & Mitchell, H. E. Role and justice considera-
tions in the attribution of responsibility to a rape victim. Journal of Researcl7 in
Personality, 1976, 10, 346-357.
U. S. Merit Systems Protection Board, Office of Merit Systems Review and Studies. Sexual
horassnient zyxwvuts
in thefederal workplace: Is it oproble~ii?
March 1981.
Volgelniann-Sine, S., Ervin, E. D., Christensen, R., Warmsun, C. H., & Ullmann, L. P.
Sex differences in feelings attributed to a woman in situations involving coercion and
sexual advances. Journal oJPersonality, 1979, 47(3),420-431.
Walster, E. Assignment of responsibility for an accident. Journal of Personaliry and Social
Weis, K., & Borges, S. S. Victiniology and rape: The case of the legitimate victim. Issues in
1978, 10(2), 196-202.
PS,YCI~OIOXY,
1966.3, 73-79.
Crin~inology,1973, 8, 71-115.