But She Was Unfaithful: Benevolent Sexism and Reactions to Rape Victims Who Violate Traditional Gender Role Expectations

Abstract

The role of benevolent sexism (BS) in accounting for victim blame in an acquaintance rape case was investigated. Participants were presented with vignettes that described an acquaintance rape. Control condition participants were given no descriptive information about the victim, whereas in the "cheating" condition the victim was described as a "married woman." As predicted, participants who scored high in BS attributed more blame to the acquaintance rape victim who was assaulted during an act of infidelity than to a victim in similar circumstances whose marital status was unknown. These findings complement those of other research (Abrams, Viki, Masser, & Bohner, in press), which indicate that individuals high in BS are more likely to react negatively to rape victims who can be viewed as violating social norms concerning appropriate conduct for women.

Full text

P1: GDP
Sex Roles [sers] PP673-sers-455296 November 8, 2002 17:41 Style file version June 3rd, 2002
Sex Roles, Vol. 47, Nos. 5/6, September 2002 ( C°2002)
Brief Report
But She Was Unfaithful: Benevolent Sexism
and Reactions to Rape Victims Who Violate
Traditional Gender Role Expectations
G. Tendayi Viki1,2 and Dominic Abrams1
The role of benevolent sexism (BS) in accounting for victim blame in an acquaintance rape
case was investigated. Participants were presented with vignettes that described an acquain-
tance rape. Control condition participants were given no descriptive information about the
victim, whereas in the “cheating” condition the victim was described as a “married woman.”
As predicted, participants who scored high in BS attributed more blame to the acquaintance
rape victim who was assaulted during an act of infidelity than to a victim in similar circum-
stances whose marital status was unknown. These findings complement those of other research
(Abrams, Viki, Masser, & Bohner, in press), which indicate that individuals high in BS are
more likely to react negatively to rape victims who can be viewed as violating social norms
concerning appropriate conduct for women.
KEY WORDS: benevolent sexism; hostile sexism; rape; victim blame.
Stereotypic beliefs about rape seem to influence
people’s evaluations of victims of sexual assault (e.g.,
Bohner et al., 1998; Krah´e, 1988). Rape myths, which
have been defined as “stereotypical beliefs about
rape that put women at a disadvantage” (Bohner &
Schwarz, 1996, p. 163), are an important example of
such beliefs. Individuals high in rape myth acceptance
(RMA) have been found to be less likely to define a
situation as a “rape” even when it meets the legally
accepted criteria (e.g., Fisher, 1986) and to attribute
more blame to the victim of rape and less blame to the
assailant (Krah´e, 1988). An example of a rape myth
is the commonly held belief that only certain types of
women (e.g., sex-workers) are usually raped (Burt,
1980). Although such a myth is clearly empirically
false, it serves the function of obscuring and denying
the personal vulnerability of all women by suggesting
that only certain kinds of women are vulnerable to
sexual violence (Lonsway & Fitzgerald, 1994).
1University of Kent at Canterbury, Canterbury, United Kingdom.
2To whom correspondence should be addressed at Department of
Psychology, University of Kent at Canterbury, Canterbury CT2
7NP, United Kingdom; e-mail: gtv1@ukc.ac.uk.
In the present research we examine victim blame
in relation to different types of victims of acquain-
tance rape. In predicting victim blame, we consider
the role of rape myth acceptance, but our main focus
is on the role of benevolent sexist beliefs.
Benevolent Sexism and Victim Blame
Glick and Fiske (1996) proposed that sexist at-
titudes may not entirely manifest in hostile forms.
Rather, sexist attitudes may be ambivalent (ambiva-
lent sexism), composed of both hostile sexism (HS)
and benevolent sexism (BS). Hostile sexism can be de-
fined as the typical antipathy that is often assumed to
characterize sexist prejudices. Benevolent sexism, on
the other hand, is a set of attitudes that are sexist but
subjectively positive and affectionate toward women.
Glick and Fiske (1996) developed the Ambivalent
Sexism Inventory (ASI), which is a 22-item measure
that assesses an individual’s level of ambivalent sex-
ism. Although, HS and BS subscales were found to be
positively correlated (Masser & Abrams, 1999), Glick
and Fiske (1996) maintained that BS and HS “have
289 0360-0025/02/0900-0289/0 C
°2002 Plenum Publishing Corporation

P1: GDP
Sex Roles [sers] PP673-sers-455296 November 8, 2002 17:41 Style file version June 3rd, 2002
290 Viki and Abrams
opposing evaluative implications, fulfilling the literal
meaning of ambivalence” (p. 494). They argued that
ambivalent sexists reconcile their hostile and benev-
olent feelings by classifying women into “good” (de-
serving BS) and “bad” (deserving HS) subcategories.
There appears to be some match between Glick and
Fiske’s theory and the existence of rape myths that
suggest that only “bad girls” are sexually assaulted.
It seems that the differentiation between “good” and
“bad” women may provide grounds for individuals to
blame certain victims of rapes.
We have previously proposed that BS may
play an important role in victim blame, especially in
acquaintance rape situations (Abrams, Viki, Masser,
& Bohner, in press). Our reasoning was based,
in part, on previous findings that women who are
perceived to have “bad reputations” are more likely
to be blamed for the occurrence of an acquain-
tance rape (Cassidy & Hurrell, 1995; L’Armand &
Pepitone, 1982; Luginbuhl & Mulin, 1981). Indeed,
lay-persons and legal practitioners have been found
to attribute blame to rape victims on the basis of
extra-legal factors such as alcohol consumption
(Corcoran & Thomas, 1991), and type of clothing
(Johnson, 1995). Such differences in victim blame
appear to be consistent with the notion that rape
victims are differentiated into “good” and “bad”
subtypes. Benevolent sexism is significantly related to
the idealization of women who conform to traditional
norms (Glick et al., 2000). Therefore, it seems likely
that individuals who are high in BS would be more
likely to be “offended” by, and hence to blame, rape
victims who may be perceived as violating traditional
gender norms.
Consistent with this argument, Abrams et al. (in
press) found that individuals high in BS assigned more
blame than low BS individuals to the acquaintance
rape victim. No differences in blame between high
and low BS individuals were observed for the stranger
rape. It is interesting that these effects were obtained
even after the effects of HS and RMA were accounted
for, which suggests that the above effects are unique
to BS. Abrams et al. argued that their findings were
obtained because high BS individuals hold particular
beliefs about how “good” women should behave and
thus view women who violate such expectations as
deserving anything unfortunate that may happen to
them.
An important limitation of Abrams et al.’s re-
search is that they use the same two vignettes (stran-
ger rape vs. acquaintance rape) across their stud-
ies. As such, it is possible that their observed re-
sults may be due to unintended subtle differences be-
tween the stimuli rather than to the general effects
they hypothesized. In this study, we focused on ac-
quaintance rape and varied the type of victim. Par-
ticipants were presented with either one of two rape
victims. For the first (control) victim, no information
concerning her personal characteristics was provided,
whereas the second victim was described as a “mar-
ried mother of three.” In essence, the description of
a married mother who is raped by an acquaintance
resulted in a scenario where a victim is sexually as-
saulted during an act that is potentially one of infi-
delity. If it is the case, as Abrams et al. argued, that
individuals high in BS blame rape victims who vio-
late traditional gender-role expectations, then a mar-
ried woman who is raped while cheating on her hus-
band should elicit very little sympathy. As such, we
expected participants to attribute more blame to the
“married mother” in comparison to the “control” vic-
tim. However, we hypothesized that this effect would
be moderated by BS such that individuals high (rather
than low) in BS would attribute more blame to the
“married mother” than to the control victim. As in
Abrams et al.’s studies, HS and RMA were not ex-
pected to moderate the effects of victim type on victim
blame.
METHOD
Participants
Fifty-seven students (29 men, 28 women) from
the University of Kent took part in this study. Partic-
ipants’ ages ranged from 18 to 39 years; 90% of the
participants were younger than 27 years (M=22.54,
SD =3.97); 78.2% of the participants were classified
as European, 14.5% as Asian, 5.5% as African, and
1.8% as “Other.”
Design
The study employed a between-subjects design,
with victim type (“control” vs. “married mother”) as
the independent variable and victim blame as the
dependent variable. Participants were randomly as-
signed to conditions. Both vignettes in this study de-
scribed a story of a woman who went to a party where
she met and became acquainted with a man. Later
that night she invited him to her apartment where, af-
ter kissing him, she was subsequently raped. However,

P1: GDP
Sex Roles [sers] PP673-sers-455296 November 8, 2002 17:41 Style file version June 3rd, 2002
Benevolent Sexism and Reactions to Rape Victims 291
the victim in one of the conditions was described as a
“married mother of three,” in comparison to the “con-
trol” condition where no details about the victim’s
characteristics were provided. The above vignettes
were identical except for the manipulation of the vic-
tim’s personal characteristics.
Materials
Seven items measured the extent to which partic-
ipants held the victim responsible for the event (e.g.,
“How much do you think Kathy should blame her-
self for what happened?”). These items were aver-
aged to provide a victim blame score for each partici-
pant (α=.74). Participants also completed the Rape
Myths Acceptance Scale (R Scale; Costin, 1985; α=
.82), the Ambivalent Sexism Inventory (Glick & Fiske,
1996; HS, α=88; BS, α=.85), and the Impression
Management Scale (IM; Paulhus, 1991; α=.60). In-
dividual differences in impression management were
measured so that this variable could be used as a so-
cial desirability control in the data analyses. A 7-point
Likert scale accompanied all the above measures.
Procedure
All the participants were approached while they
were studying in the university library and asked
whether they would complete a questionnaire on
“gender relations.” Those individuals who volun-
teered to participate were handed a questionnaire
that contained either the “control” or “married
mother” vignette and were left to complete the ques-
tionnaire on their own. In the questionnaire, partici-
pants first read the scenario that depicted the rape and
then responded to the questions about victim blame.
After this, participants responded to the ASI, Costin’s
R, and the IM scale, respectively. When the question-
naires had been completed, the researcher returned
to thank and debrief the participants and to collect
the questionnaires.
RESULTS
Preliminary analyses showed that gender did not
reveal any significant main or interaction effects (with
BS, HS, or RMA) on victim blame (all ps>.05).
Therefore, gender is not discussed in further anal-
yses. Correlation analyses were performed across
Table I. Correlations Among Measures of Rape Myth Acceptance,
Hostile Sexism, Benevolent Sexism, and Victim Blame
Benevolent Hostile Rape myth
sexism sexism acceptance
Benevolent sexism —
Hostile sexism .50 —
Rape myth acceptance .42 .52 —
Victim blame .32 .43 .57
Note. All correlations significant at p<.05.
conditions and this yielded significant correlations
among all the measures (all ps<.01; see Table I).
To analyze the impact of victim type and BS, hi-
erarchical regression analysis was employed. In the
first step, victim type and BS were entered, and in the
second step the interaction term (BS ×Victim Type)
was entered. The overall equation was significant,
F=6.81, p<.01. Significant main effects for victim
type and BS were obtained (see beta coefficients in
Table II). As expected, more blame was attributed
to the “married mother” than to the “control” victim
(M=3.78, SD =1.13; M=3.03, SD =0.81, respec-
tively). The significant positive relationship between
BS and victim blame indicates that the higher individ-
uals scored on BS, the more they blamed the rape vic-
tim. These main effects were qualified by a significant
interaction between BS and victim type ( p=.047; see
Table II). Simple effects analyses revealed that in the
control condition the relationship between BS and
victim blame was positive but failed to reach signifi-
cance, β=.14, t(31) =0.77, ns. Although this result
is in the expected direction, it is not consistent with the
results from our previous research (e.g., Abrams et al.,
in press), where the positive relationship between BS
and victim blame was found to reach statistical sig-
nificance in a similar condition. In contrast, in the
“married mother” condition there was a significant
relationship between victim blame and BS, β=.51,
t(24) =2.84, p<.01.Thus, the higher individuals
scored on BS, the more they blamed the “married
mother” for the rape. These effects remained signifi-
cant even after the effects of IM, HS, and RMA were
Table II. Regression Analysis of the Effects of Benevolent Sexism
and Victim Type on Victim Blame
Regression step βTSig. rprsrR
2
change
Step 1
Benevolent sexism .30 2.44 .018 .32 .32 .29
Victim type .35 2.88 .006 .37 .37 .35 .22
Step 2
BS ×Victim Type .30 2.03 .047 .34 .27 .24 .06

P1: GDP
Sex Roles [sers] PP673-sers-455296 November 8, 2002 17:41 Style file version June 3rd, 2002
292 Viki and Abrams
partialled out before entering BS, victim type, and the
interaction term in the regression, pr =.29, β=.25,
t(56) =2.00, p=.05.Separate hierarchical regres-
sion analyses (similar to those conducted for BS) were
performed to examine whether HS and RMA inter-
acted with victim type to predict victim blame. As pre-
dicted, the interaction effects between victim type and
HS or RMA failed to reach significance (all ps>.05).
DISCUSSION
The present results are consistent with our main
hypotheses. As expected, participants assigned more
blame to the adulterous “married mother” in com-
parison to the “control” victim. This is consistent with
previous research (e.g., Cassidy & Hurrell, 1995) and
suggests that perceptions of the appropriateness of
the victim’s behavior may have some influence on the
participants’ reactions to different types of acquain-
tance rape victims. This proposal is further supported
by the finding that individuals who are high in BS
attributed more blame to the “married mother” than
did low BS individuals. Thus, BS appears to provide an
alternative mechanism through which some of the ob-
served differences in victim blame can be accounted
for. According to Glick et al. (2000), benevolent sex-
ists have a tendency to idealize women who conform
to traditional gender roles As such, a married woman
who is raped during an act of potential infidelity could
be viewed as grossly violating high BS individuals’ be-
havioral standards and may then be blamed for the
event.
Although several researchers have reported that
rape victims who violate traditional gender roles are
more likely to be blamed (e.g., Cassidy & Hurrell,
1995; Krah´e, 1988), the present results are interesting
because they suggest that BS is an important mod-
erator of previously reported findings. Furthermore,
previous researchers have tended to focus on the role
of sexist hostility and adversarial sexual beliefs in the
evaluation of rape victims (e.g., Krah´e, 1988). In fact,
Lonsway and Fitzgerald (1995) noted that RMA is
mainly driven by hostile attitudes toward women. The
present study adds to the research in this area by show-
ing that subjectively positive sexist attitudes (BS) may
also play an important role in the negative evaluations
of some rape victims. Indeed, RMA did not moder-
ate the effect of victim type on victim blame after the
effects of BS were accounted for.
In the present research, we studied a student sam-
ple using hypothetical scenarios and paper and pencil
responses. Therefore, future studies are required to
explore whether the findings reported here can be
replicated in more applied settings (e.g., jury decision
making) and with different types of participants and
cultures. Benevolently sexist attitudes often go un-
challenged in broader society because they are per-
ceived as prosocial (Glick & Fiske, 1996). However,
the present research demonstrates some potentially
negative effects that could have serious consequences
for the treatment and evaluation of rape victims.
ACKNOWLEDGMENT
Preparation of this paper was made possible by a
postgraduate research grant F3-37A-2 to G. Tendayi
Viki by the Beit Trust, UK.
REFERENCES
Abrams, D., Viki, G. T., Masser, B., & Bohner, G. (in press). Per-
ceptions of stranger and acquaintance and stranger rape: The
role of benevolent and hostile sexism in victim blame and rape
proclivity. Journal of Personality and Social Psychology.
Bohner, G., Reinhard, M., Rutz, S., Sturm, S., Kerschbaum, B., &
Effler, D. (1998). Rape myths as neutralizing cognitions: Ev-
idence for a causal impact of anti-victim attitudes on men’s
self-reported likelihood of raping. European Journal of Social
Psychology,28, 257–268.
Bohner, G., & Schwarz, N. (1996). The threat of rape: Its psycho-
logical impact on non-victimized women. In D. M. Buss & N.
Malamuth (Eds.), Sex, power, conflict: Evolutionary and femi-
nist perspectives (pp. 162–175). New York: Oxford University
Press.
Burt, M. R. (1980). Cultural myths and supports of rape. Journal of
Personality and Social Psychology,38, 217–230.
Cassidy, L., & Hurell, R. M. (1995). The influence of victim’s attire
on adolescents’ judgments of date rape. Adolescence,30, 319–
324.
Corcoran, K. J., & Thomas, L. R. (1991). The influence of observed
alcohol consumption on perceptions of initiation of sexual ac-
tivity in a college dating situation. Journal of Applied Social
Psychology,21, 500–507.
Costin, F. (1985). Beliefs about rape and women’s social roles.
Archives of Sexual Behavior,14, 319–325.
Fischer, G. J. (1986). College student attitudes toward forcible rape:
I. Cognitive predictors. Archives of Sexual Behavior,15, 457–
466.
Glick, P., & Fiske, S. T. (1996). The Ambivalent Sexism Inven-
tory: Differentiating hostile and benevolent sexism. Journal
of Personality and Social Psychology,70, 491–512.
Glick, P., Fiske, S. T., Mlandinic, A., Saiz, J. L., Abrams, D., Masser,
B., et al. (2000). Beyond prejudice as a simple antipathy:
Hostile and benevolent sexism across cultures. Journal of
Personality and Social Psychology,79, 763–775.
Johnson, K. K. P. (1995). Attributions about date rape: Impact of
clothing, sex, money spent, date type, and perceived similarity.
Family and Consumer Sciences Research Journal,23, 292–311.
Krah´e, B. (1988). Victim and observer characteristics as determi-
nants of responsibility attributions to victims of rape. Journal
of Applied Social Psychology,18, 50–58.

P1: GDP
Sex Roles [sers] PP673-sers-455296 November 8, 2002 17:41 Style file version June 3rd, 2002
Benevolent Sexism and Reactions to Rape Victims 293
L’Armand, K., & Pepitone, A. (1982). Judgments of rape: A study
of victim–rapist relationship and victim sexual history. Person-
ality and Social Psychology Bulletin,8, 134–139.
Lonsway, K. A., & Fitzgerald, L. F. (1994). Rape myths: In review.
Psychology of Women Quarterly,18, 133–164.
Luginbuhl, J., & Mulin, C. (1981). Rape and responsibility: How
and how much is the victim blamed? Sex Roles,7, 547–559.
Masser, B., & Abrams, D. (1999). Contemporary sexism: The
relationships among hostility, benevolence, and neosexism.
Psychology of Women Quarterly,23, 503–517.
Paulhus, D. L. (1991). Measurement and control of response bias.
In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.),
Measures of personality and social psychological attitudes
(pp. 17–59). San Diego: Academic Press.