The Journal of Social Psychology, 2002, 142(2), 264271

Application of the AttributionValue Model

of Prejudice to Homosexuality
NURAY SAKALLI
Psychology Department
Middle East Technical University (METU), Ankara, Turkey

ABSTRACT. According to the attributionvalue model, prejudice toward a group stems

from 2 interrelated variables: attributions of controllability and cultural value. Thus, prejudice toward gay men and lesbians may stem from others holding them responsible for
their behaviors and perceiving negative cultural value regarding homosexuality. The
author tested that model by using the issue of homosexuality. The participants were Turkish undergraduates who completed a homophobia scale and answered questions about the
origins of homosexuality, cultural attitudes toward homosexuality, and their own gender
and sexual preferences. In general, the participants were prejudiced against gay men and
lesbians. As expected, attributions of controllability and negative cultural value regarding
homosexuality operated jointly to explain 39% of the variation in homophobia. The participants who thought that homosexuality was controllable had more negative attitudes
toward gay men and lesbians than did those who thought that homosexuality was uncontrollable. The female participants were more tolerant of homosexuality than were the male
participants.
Key words: attitudes, attributionvalue model, controllability, cultural value, gender differences, homosexuality, gay men, lesbians, prejudice

PREJUDICE has been an important research topic in social psychology.

Researchers have explored variables that lead to prejudice. Crandall and his
coworkers (Crandall et al., 2001; Crandall & Martinez, 1996) have developed the
attributionvalue model of prejudice: Prejudice toward groups is based on two
interrelated variablesattributions of controllability and cultural value. Focusing
on antifat antipathy, Crandall et al. suggested that attributions and negative value
regarding fatness separately predict antifat prejudice. However, the simultaneous
presence of attributions and negative cultural value was more likely to predict
antifat prejudice. Thus, people who both made attributions of controllability and
devalued fatness were more likely to express antifat prejudice.
Address correspondence to Nuray Sakalli, Psychology Department (Psikoloji Blm),
Middle East Technical University (Orta Dogu Teknik niversitesi), 06531, Ankara,
Turkey; nurays@metu.edu.tr (e-mail).
264

Sakalli

265

In the present study, I focused on the relationship between heterosexual participants causal attributions for the origins of homosexuality and their attitudes
toward homosexuality. I especially wanted to test the attributionvalue model of
prejudice by focusing on beliefs, values, and prejudice with respect to gay men
and lesbians. Examinations of the attributionvalue model of prejudice with
groups other than overweight people (Crandall et al., 2001) may broaden the conceptions of prejudice and show the applicability of the new model of prejudice to
other stigmatized groups.
Attribution theories deal with how social perceivers use information to arrive
at causal explanations for eventsthat is, how individuals use certain cues or
dimensions to form a causal judgment. According to the attribution model (Weiner, 1995), for example, the dimensions of locus, stability, and controllability
determine the emotional experiences of anger, guilt, gratitude, pity, pride, hopelessness, and shame. That model suggests that positive events attributed to controllable behavior elicit positive affect, whereas negative events attributed to controllable behavior elicit negative affect: If a person experiences a negative or
undesirable outcome and the cause is controllable, then the observer may hold
that person responsible for the negative outcome and, consequently, may feel
negative affect (e.g., anger) toward the person. However, if there is an undesirable outcome and the cause is uncontrollable, then the observer may experience
pity and offer help to the person involved (Weiner, 1985).
Attributions of controllability stem from underlying beliefs about causality in
the physical and social world. Lerner (1980) suggested that one who believes that
people get what they deserve (i.e., behave in such a way as to cause a particular
outcome) also treats such people in accordance with the value of their outcomes.
Thus, one would punish, avoid, and stigmatize people with negative characteristics or outcomes (e.g., overweight, poverty, AIDS, homosexuality, physical handicaps) because such people deserve the anger and prejudice of others (Feather,
1996). For example, previous researchers (Allison, Basile, & Yuker, 1991; Crandall, 1994) have shown a clear association between judgments of responsibility for
excess weight and negative reactions to overweight people. People are more willing to derogate a person who is overweight from overeating (controllable) than one
who is overweight from a glandular disorder (uncontrollable; DeJong, 1980).
According to the attributionvalue model of prejudice (Crandall et al.,
2001), attributions of controllability are not the only causes of prejudice. In
examinations of prejudice, one must also consider the interrelationship between
attributions of controllability and cultural value. Because prejudice is defined as
negative affect toward members of groups, prejudice against gay men and lesbians may derive not only from holding gay men and lesbians responsible for
their attributes but also from negative cultural attitudes toward them. The development of negative attitudes toward gay men and lesbians requires that the attributes of gay and lesbian individuals be perceived negatively in a society. To better understand prejudice against gay men and lesbians, I included the issue of

266

The Journal of Social Psychology

negative cultural value in the study, as suggested by the attributionvalue model

of prejudice (Crandall et al.).
Application of the attributionvalue model to homosexuality is important as
evidence that the model is applicable to stigmatized groups other than overweight people. Following the model, I have suggested that heterosexual individuals who believe that homosexuality is learned rather than genetic hold gay and
lesbian individuals responsible for their life style. Those heterosexual individuals may perceive that gay men and lesbians choose their lifestyle and have control over their sexual preferences. As Whitley (1990) suggested, the perception
that homosexuality is a preference may lead to negative attitudes among heterosexual individuals. However, if heterosexual individuals perceive that homosexuality has genetic causes and that gay men and lesbians, therefore, cannot control their sexual preferences, they may not dislike or reject them as much. In
addition, Aguero, Bloch, and Byrne (1984) suggested that the belief that homosexuality is genetic produces attitudes toward gay men and lesbians basically
similar to attitudes toward the handicapped because, in both cases, people do not
perceive genetically determined behaviors as controllable. The undesirable position of gay men and lesbians and the perception that the cause of homosexuality
is uncontrollable may lead to pity (Weiner, 1985) and less negative attitudes
toward their life style.
In short, the purpose of the present study was to test the attributionvalue
model of prejudice in the case of homosexuality. In accord with the preceding literature on homosexuality and prejudice, I formulated the following hypotheses:
Hypothesis 1: The simultaneous presence of an attribution of controllability
and negative cultural value is most likely to predict negative attitudes toward gay
men and lesbians.
Hypothesis 2: The participants who believe that homosexuality is controllable are more prejudiced against gay men and lesbians than are those who
believe that homosexuality is uncontrollable.
Hypothesis 3: In general, the participants hold negative attitudes toward gay
men and lesbians. Consistent with earlier findings (Kite, 1984; Sakalli, 2002), the
female participants are less prejudiced toward gay men and lesbians than are the
male participants.

Method
Participants and Procedure
The participants (age range = 1726 years; M = 20.64, SD = 1.94) were 307
undergraduates (117 female and 190 male) at Middle East Technical University
(METU) in Ankara, Turkey. Engineering and business students attending elective
introductory courses in social psychology and psychology filled out a question-

Sakalli

267

naire in a classroom situation. I thanked them and gave them course credit for
their participation.
The participants were from various regions of Turkey (data were missing for
1 participant)the Sea of Marmara (60 participants), the Mediterranean Sea
(48), The Black Sea (26), the Aegean Sea (41), Inner Anatolia (117), and Southeast Anatolia and East Anatolia (14). Even though many of the participants were
from big citiesfor example, Ankara (95), Istanbul (24), Izmir (19), Izmit (19),
Adana (16), and Bursa (15)there were also participants from various other
cities (63 indicated by the participants; total N in Turkey = 84).
Instruments
Attitudes toward gay men and lesbians. I measured the participants attitudes
toward gay men and lesbians by using the Homophobia Scale (Hudson &
Ricketts, 1980). Two Turkish students whose second language was English separately translated the scale from English into Turkish. Then, a teacher in the
department of modern languages at METU translated the Turkish version of the
scale back into English.
To assess their attitudes toward gay men and lesbians, the participants indicated their agreement or disagreement with each item (e.g., I would feel nervous
being in a group of homosexuals, It would disturb me to find out that my doctor was homosexual) by using a 6-point Likert-type scale (1 = agreement
negative attitudes, 6 = disagreementpositive attitudes).
Attributions of controllability. I measured attributions of controllability by asking the participants to indicate on a 6-point Likert-type scale their disagreement
or agreement (1 = disagreement, 6 = agreement) with the statement I believe
that homosexuality is due to learning and preference and not to biology and
genetics. Then, for the analysis of variance (ANOVA), I classified the participants who circled 1, 2, and 3 as believing that homosexuality was uncontrollable
and those who circled 4, 5, and 6 as believing that homosexuality was uncontrollable.
Cultural Value
I measured cultural value regarding gay men and lesbians by asking the participants first to think about how Turkish society, in general, perceived homosexuality and then to indicate on a 6-point Likert-type scale disagreement or
agreement (1 = disagreementpositive cultural value, 6 = agreementnegative
cultural value) with the statement Our society does not accept homosexuality
and gay men and lesbians. Last, the participants indicated their gender and sexual preferences (all heterosexual) on the questionnaire.

268

The Journal of Social Psychology

Results
Attitudes Toward Homosexuals
A principal components factor analysis of the items of Hudson and Rickettss
(1980) scale, with varimax rotation of factors with eigenvalues greater than 1.0,
indicated that three dimensions described the data: The factors were Social Contact With Gay Men and Lesbians (Factor 1; = .92), Probable Family Ties With
Gay Men and Lesbians (Factor 2; = .40), and Tendency to be a Gay Man/Lesbian (Factor 3; = .62). Factor 1 (eigenvalue = 10.30) accounted for 26.77% of
the variance, Factor 2 (eigenvalue = 1.73) accounted for 17.07% of the variance,
and Factor 3 (eigenvalue = 1.24) accounted for 11.49% of the variance. The whole
scale accounted for 55.35% of the total variance (for the whole scale, = .93).
The mean score for the entire Homophobia Scale (Hudson & Ricketts, 1980)
was 2.75 (SD = 1.02), and the mean scores for the factors Social Contact With Gay
Men and Lesbians, Probable Family Ties With Gay Men or Lesbians, and Tendency
to be a Gay Man/Lesbian were 3.30 (SD = 1.31), 2.28 (SD = 1.03), and 2.27 (SD =
1.06), respectively. Thus, the participants attitudes toward gay men and lesbians
were on the negative side given that the scale ranged from 1 to 6, with higher scores
indicating more positive attitudes toward gay men and lesbians. The most positive
attitude was related to the factor Social Contact With Gay Men and Lesbians.
The Relationship Between Gender Differences, Attributions of Controllability,
and Attitudes Toward Gay Men and Lesbians
Of the participants, 164 indicated that homosexuality was uncontrollable;
128 indicated that it was controllable; and 15 did not answer.
To explore the relationship between gender differences, attributions of controllability, and attitudes toward homosexuals, I performed a 2 (gender of respondent) 2 (attributions of controllability) ANOVA on attitudes toward gay men and
lesbians. The results of the analysis demonstrated that there were main effects for
gender, F(1, 291) = 4.86, p < .05, and attributions of controllability, F(1, 291) =
70.79, p < .01, but no significant interaction between gender differences and attributions of controllability, F(1, 291) = .55, p = .45. The female participants (M =
3.01, SD = 1.03) had more positive attitudes toward gay men and lesbians than did
the male participants (M = 2.59, SD = 0.99). The participants who thought that
homosexuality was controllable (M = 2.20, SD = 0.85) had more prejudiced attitudes toward gay men and lesbians than did those who thought that homosexuality was uncontrollable (M = 3.17, SD = 0.95). Regression analysis demonstrated
that 21% of the variation in attitudes toward gay men and lesbians was explained
by attributions of controllability, R2 = .21, F(1, 290) = 80.89, p < .01.
Prejudice, Attributions of Controllability, and Cultural Value
To test the effects of attributions of controllability and cultural value on prej-

Sakalli

269

udice, I correlated attitudes toward gay men and lesbians with cultural values and
attributions of controllability. Attitudes toward gay men and lesbians were significantly correlated with both attributions of controllability, r(292) = .46, p <
.01, and cultural value, r(292) = .54, p < .01. Thus, the participants negative
attitudes toward gay men and lesbians increased when they believed that homosexuality was learned and when they perceived negative cultural value.
According to regression analysis, 39% of the variation in attitudes toward
gay men and lesbians was explained by linear regression on the attributions of
controllability and negative cultural value operating jointly, R2 = .39, F(2, 289) =
94.21, p < .01. Both attributions of controllability, = .34, p < .01, and cultural values, = .44, p < .01, significantly influenced the participants attitudes
toward gay men and lesbians.
Discussion
I conducted the present study to test the attributionvalue model of prejudice
(Crandall et al., 2001). The findings demonstrated that attitudes in Turkey toward
gay men and lesbians were explained by attributions of controllability and cultural values with respect to homosexuality. The results supported the idea that the
causes of prejudice may be the simultaneous presence of an attribution of controllability and a negative cultural value with respect to out-groups. Both attributions of controllability and cultural value regarding homosexuality were related
to prejudiced attitudes toward gay men and lesbians.
Consistent with earlier studies (Aguero et al., 1984; Whitley, 1990), the present study demonstrated that, when the participants perceived the origin of homosexuality as controllable, they were more prejudiced toward gay men and lesbians.
Thus, controllable causes of homosexuality may have led the present heterosexual participants to hold gay men and women responsible for their own life styles
and, consequently, to express negative attitudes toward homosexuality. That finding generally supported attribution theory (Weiner, 1985). As expected, the more
controllable the participants perceived the causes of the stigma, the more anger
and prejudice they directed toward the stigmatized group (Feather, 1996).
Whitley (1990) suggested that the relationship between perceived controllability of the origins of homosexuality and antigay attitudes among heterosexuals
explains how some college sexuality courses successfully ameliorate such attitudes. Whitley (1990) argued that textbooks in that area usually present a number of biological bases for homosexuality, implying uncontrollability. Thus, if
individuals believe that the origin of homosexuality is not learned but, rather, biological (i.e., cannot be changed), then they may not hold gay men and lesbians
responsible for their sexual preferences and, therefore, may hold less negative
attitudes toward homosexuality.
The present results also support the idea that negative cultural value is an
important predictor of prejudice against gay men and lesbians. Earlier

270

The Journal of Social Psychology

researchers have used gender, race, and excess weight to examine prejudice (e.g.,
Crandall et al., 2001). However, the present study is the first to focus on the issue
of homosexuality to test the attributionvalue model of prejudice (Crandall et
al.). My results demonstrated that the addition of negative cultural value to the
causes of prejudice against homosexuality increased the percentage of variation
in antihomosexual prejudice in regression analysis, suggesting that both negative
cultural value and attribution of controllability lead heterosexual individuals to
hold prejudiced attitudes toward gay men and lesbians.
Consistent with earlier findings (e.g., Kite, 1984), the present finding about
gender differences demonstrated that the present male participants were more
prejudiced toward homosexuals than were the present female participants. As
many researchers have argued, male participants may be more negative toward
homosexuals than are female participants because of (a) their attitudes concerning paternalistic power and gender differentiation and (b) their tendency to be
less tolerant of gay men (Milham, San Miguel, & Kellogg, 1976; Whitley, 1987).
Researchers (Black & Stevenson, 1984; Sakalli, 1999) have found that many
individuals thought of men when they were given the concept of homosexuality.
The perception that homosexuals are usually men may have led the present male
participants to fear gay men and the possibility of being the targets of male
homosexual advances. The present female participants may have had less negative attitudes toward homosexuality because the scale (Hudson & Ricketts, 1980)
did not specify the gender of the homosexual target; therefore, they may have
thought of men as the targets in their responses to statements about homosexuals. The female participants may not have felt uncomfortable because they did
not fear being the targets of male homosexual advances.
The present study contributes to the literature by replicating the results of
earlier studies about the effects of attributions of causality on attitudes toward
homosexuality (Whitley, 1990). By using homosexuality, a different criterion
from those in earlier studies (e.g., overweight; Crandall et al., 2001), it also
demonstrated that the attributionvalue model of prejudice was a good model to
explain what variables lead to prejudice in a collectivist country.
However, there are some limitations of the present study. Because it was a
correlational study, I could not draw concrete conclusions for the support of the
attributionvalue model. Future researchers should conduct experimental designs
to find a causal relationship between attributions of controllability, cultural value,
and antipathy toward out-groups.
In addition, I used only undergraduates at METU. As indicated earlier, the
present students were from different regions of Turkey; such distribution suggests
that my sample was a good general representation of Turkish people. However,
several studies in Turkey have demonstrated that the levels of students sexism,
conservatism (Sakalli, 2002), and social contact with gay men and lesbians
(Sakalli & Ugurlu, 2001) may influence the negativity or positivity of attitudes
toward them. Thus, it may be appropriate to use nonstudent participants, who may

Sakalli

271

have different levels of sexism and conservatism from those of students, to

increase the external validity of the present study. Furthermore, researchers should
conduct more cross-cultural studies to examine whether the attributionvalue
model of prejudice is supported in other cultures for different stigmatized groups.
REFERENCES
Aguero, J. E., Bloch, L., & Byrne, D. (1984). The relationships among sexual beliefs, attitudes, experience, and homophobia. Journal of Homosexuality, 10, 95107.
Allison, D. B., Basile, V. C., & Yuker, H. E. (1991). The measurement of attitudes toward
and beliefs about obese persons. International Journal of Eating Disorders, 5, 599607.
Black, K. N., & Stevenson, M. R. (1984). The relationship of self-reported sex-role characteristics and attitudes toward homosexuality. Journal of Homosexuality, 10, 8393.
Crandall, C. S. (1994). Prejudice against fat people: Ideology and self-interest. Journal of
Personality and Social Psychology, 66, 882894.
Crandall, C. S., DAnello, S., Sakalli, N., Lazarus, E., Nejtardt, G. W., & Feather, N. T.
(2001). An attributionvalue model of prejudice: Anti-fat attitudes in six nations. Personality and Social Psychology Bulletin, 27, 3037.
Crandall, C. S., & Martinez, R. (1996). Culture, ideology, and anti-fat attitudes. Personality and Social Psychology Bulletin, 22, 11651176.
DeJong, W. (1980). The stigma of obesity: The consequences of nave assumptions concerning causes of physical deviance. Journal of Health and Social Behavior, 21, 7587.
Feather, N. T. (1996). Reactions to penalties for an offense in relation to authoritarianism,
values, perceived responsibility, perceived seriousness, and deservingness. Journal of
Personality and Social Psychology, 71, 571587.
Hudson, W., & Ricketts, W. (1980). A strategy for measurement of homophobia. Journal
of Homosexuality, 5, 357372.
Kite, M. E. (1984). Sex differences in attitudes toward homosexuals: A meta-analytic
review. Journal of Homosexuality, 10, 6981.
Milham, J., San Miguel, C. L., & Kellogg, R. (1976). A factor-analytic conception of attitudes toward male and female homosexuals. Journal of Homosexuality, 2, 310.
Lerner, M. J. (1980). The belief in a just world: A fundamental delusion. New York: Plenum.
Sakalli, N. (1999). [The perception about the gender of homosexuals]. Unpublished raw
data.
Sakalli, N. (2002). The relationship between sexism and attitudes toward homosexuality
in a sample of Turkish College students. Journal of Homosexuality, 42(3), 5162.
Sakalli, N., & Ugurlu, O. (2001). Effects of social contact with homosexuals on heterosexual Turkish university students attitudes toward homosexuality. Journal of Homosexuality, 42(1), 5362.
Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92, 548573.
Weiner, B. (1995). Judgments of responsibility. New York: Guilford.
Whitley, B. E. (1987). The relationship of sex-role orientation to heterosexuals attitudes
toward homosexuals. Sex Roles, 17, 103113.
Whitley, B. E. (1990). The relationship of heterosexuals attributions for the causes of
homosexuality to attitudes toward lesbians and gay men. Personality and Social Psychology Bulletin, 16, 369377.

Received March 27, 2000

Accepted December 8, 2000