University of Southern California Dissertations and Theses

Heterosexuals' attitudes toward lesbians and gay men and the attitude functions they serve: Correlates, stability and gender differences

(USC Thesis Other)

Heterosexuals' attitudes toward lesbians and gay men and the attitude functions they serve: Correlates, stability and gender differences

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content HETEROSEXUALS’ ATTITUDES TOWARD LESBIANS AND GAY MEN AND
THE ATTITUDE FUNCTIONS THEY SERVE:
CORRELATES, STABILITY AND GENDER DIFFERENCES
by
Julie Lynn Anderson
A Dissertation presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(EDUCATION-COUNSELING PSYCHOLOGY)
December 2003
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3133241
Copyright 2004 by
Anderson, Julie Lynn
All rights reserved.
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 3133241
Copyright 2004 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90089-1695
This dissertation, written by
-J 0 t. X& L - Y H Nj -
under the direction o f k & £ _ dissertation committee, and
approved by all its members, has been presented to and
accepted by the D irector o f Graduate and Professional
Programs, in partial fulfillment of the requirements fo r the
degree of
DOCTOR OF PHILOSOPHY
Director
Date_ _ D e c e a ih £ ] L _ ll^ M B
Dissertation Committee
7
Chair
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
I would like to thank my faculty advisor, Dr. Ruth Chung for providing direction,
guidance and support throughout the entire dissertation process. I am grateful for the
inspiring and encouraging conversations I have had with Dr. Walter Williams, who also
served on my dissertation committee. I also would like to thank Dr. Rodney Goodyear
for being the most consistent academic connection I had throughout my schooling at USC
and for taking the time to help me analyze and interpret the data for this project.
I am grateful to Dr. Christine Browning who was not only a clinical mentor to me,
but who also introduced me to Dr. Gregory Herek whose advice I appreciated and whose
work I respect and inspired me to conduct this study. I am grateful to Joan Garry and my
colleagues at GLAAD for being flexible and supportive when work and school
requirements sometimes overlapped.
My friends Mary & Mary, Tammy, Marta, Matt, Jordana & Christina, Cindy,
Susan, Julie & Kelly, Tanya, Laurie, Damley, Julia, Rhona and Deborah & Maggie
provided me with a consistent supply of support and well wishes. I would like to thank
schoolmate Dr. Chris Hodson who provided moral support, advice and friendship
throughout the entire dissertation process, not to mention a couple of great meals. And,
thanks to my friend Rachael who, many years ago, advised me to enjoy the journey.
Reaching the goal of earning a doctorate has allowed me to look back and savor the rich
and myriad memories I have accumulated along the way.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iii
I would like to thank my parents, Norman and Joan Anderson for believing in my
ability to earn my Ph.D. and for instilling in me the importance of having a strong work
ethic and to always believe in myself. I thank my sisters Laurie and Kristie, both who
have been willing to take risks to live their dreams.
I would especially like to thank my loving partner, Amy Dantzler who inspires
excellence and who was always there for me. She has been a rock and my greatest
supporter throughout my entire journey through USC.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv
TABLE OF CONTENTS
Page
Acknowledgements. ............ ii
List of Tables.. ....... vii
Abstract. ..............................................x
Chapters
I. Introduction. ...................... 1
Background of the Problem ........................ 1
Definitions of Attitude. ................ .6
Attitude Representation and Change Theory ......... ...7
Influence of context on attitudes .............. 8
Self-concept theory and attitudes............................... 11
Attitude change theory............... ....12
Tensor Product Model (TPM)............................... 15
Influence of disconfirming information about
stereotypes on stereotype c h a n g e ..............................16
Summary of attitude formation and change theory 17
Attitude Function Theory ..........................................................18
Development of the attitudes function inventory.. 20
Person, domain and situational characteristics..............24
Summary of attitude function theory.............................26
Overview of the Study. .....................26
II. Review of the Literature. ....... 28
Attitudes Toward Gay Men and Lesbians ........... 28
Sex, age, educational level, religiosity, income,
adherence to traditional ideologies of family and
gender, perception of friends’ agreement with
one’s own attitudes, contact with gay men and
lesbians and residence..................... .29
Cognitive and affective bases for attitudes toward
gay men and lesbians. .............. 33
Attitude strength and attitudes toward gay men and
lesbians ........ .35
Influence of educational strategies on attitudes
toward gay men and lesbians. ............... 36
Influence of out-group versus in-group messenger
on attitudes toward gay men and lesbians ........39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
Summary of attitudes toward gay men and lesbians... ..40
Contact Hypothesis. ............ .41
Attitude activation and exemplars ...... .42
Contact with gay men and lesbians ...... ...45
Cognitive dissonance. ....... .46
Contact hypothesis and attitudes toward gay men
and lesbians. ....... 48
Summary of influence of contact hypothesis on
attitudes toward gay men and lesbians .....52
Media Effects. ........... .53
Summary of media influences on attitudes toward
gay men and lesbians. ................. .57
Attitude Function Research................ .57
Matching versus non-matching functional
approach to attitude change...........................................63
Summary of attitude function research. ........ 67
Gender Differences and Item Order. .................. 68
Item order differences.................................. 74
Summary of gender differences and item order. .....78
Synthesis of Review of the Literature................... ....79
Purpose of the Study. ....................... 83
Research Questions and Hypotheses. ...................................84
III. Method................ 86
Participants ......... 86
Instrument. .................. ......89
Attitudes toward gay men and lesbians scales. .........90
Attitude functions scales—revised. ........91
Procedure.. ........... 93
IV. Results. ....... 96
Intercorrelations ....... 96
Research question 1: Are there demographic and social
group differences between heterosexuals’ attitudes
toward lesbians, attitudes toward gay men and the
functions these attitudes serve? ........... ........99
Research question 2: Are there sex differences in opinions
about gay and lesbian civil rights?.................................1 0 7
Research question 3: Do attitude functions predict
attitudes toward lesbians and attitudes toward gay men?.......109
Research question 4: Are there sex and item order
differences in attitudes toward lesbians, attitudes toward
gay men and attitude functions?. ....... .114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
V. Discussion
Discussion of Results
118
119
Hypothesis la: There will be sex, age, education,
religious ideology, religious attendance, religious
affiliation, media images, geographic residence
and degree of closeness with lesbians and gay men
differences for heterosexuals’ attitudes toward gay
men and heterosexuals’ attitudes toward lesbians.. ....119
Hypothesis lb: Heterosexual men will have less
favorable attitudes toward gay men and toward
lesbians than heterosexual women....................... 123
Hypothesis 1c: There will be a significant
relationship between the number of gay men and
lesbians someone is close to and the defensive
function. ............ 125
Hypothesis Id: There will be gender group
differences for the defensive function................... 126
Hypothesis 2: Heterosexual men will oppose civil
rights for gay men and lesbians more than will
heterosexual women.......................... ....................1 2 8
Hypothesis 3a: The defensive function will predict
less favorable attitudes toward lesbians and toward
gay m en ............................... ...............132
Hypothesis 3b: The social-expressive function will
predict less favorable attitudes toward lesbians and
toward gay men.................................... ......133
Hypothesis 3c: The value-expressive function will
predict less favorable attitudes toward lesbians and
toward gay men .........................134
Hypothesis 3d: The experiential-schematic function
will predict more favorable attitudes toward
lesbians and toward gay men ...... 135
Hypothesis 4a: Heterosexual men’s attitudes
toward lesbians will be more unfavorable when
questions about gay men are asked first................136
Hypothesis 4b: Heterosexual men will exhibit an
increased defensive function when questions about
gay men are asked before questions about lesbians... 139
Hypothesis 4c: Assessing attitudes about gay men
first will make the defensive function more
accessible for male respondents, resulting in more
unfavorable attitudes toward gay men. ........... ..139
Summary. ................................ 140
Implications of the Study. .....144
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii
Limitations of the Study. ......147
Future Directions.................................................... 149
Conclusion........................................................... 153
References................. ..................... ......................................................... 156
Appendix A: Attitude and Attitude Function Survey ...... .....168
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table Page
1. Profile of General Demographic Characteristics of the 2000
United States Census and Current Study.. ......................... 88
2. Zero-order Pearson Product Correlations for Demographic and
Measured Variables...............................................................9 8
3. Mean ATL and ATG Scores by Respondent Sex, Residence,
Religious Attendance, Religious Ideology, Religious Affiliation,
Closeness to Gay Men and Closeness to Lesbians............................1 0 0
4. Multivariate Analysis of Variance for ATL and ATG by Sex,
Residence, Religious Attendance, Religious Ideology, Religious
Affiliation and Closeness to Gay Men and Closeness to Lesbians........... 103
5. Multivariate Analysis of Variance for the Defensive Function
by Sex and Closeness to Gay Men and Closeness to Lesbians ..106
6. Comparison of Past Responses to Four Public Policy Questions
and Respondent Sex Breakdown for this Study.. ..................................108
7. Multivariate Analysis of Variance for Four Public Policy
Issues by Sex .................................................. ......................................... 109
8. Variance Explained by Attitude Functions Predicting Men’s
and Women’s ATL and ATG Scores. ......................................................110
9. Summary of Stepwise Regression Analysis for Men’s Attitudes
Toward Lesbians.. ................................................................. ..110
10. Summary of Stepwise Regression Analysis for Men’s Attitudes
Toward Gay Men ...... ....Ill
11. Summary of Stepwise Regression Analysis for Women’s
Attitudes Toward Lesbians.. ................. ..112
12. Summary of Stepwise Regression Analysis for Men’s Attitudes
Toward Gay M en .................................... ..113
13. Mean ATL and ATG Scores by Respondent Sex and Item Order. 114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14. Mean AF1-L and AFI-G Scores by Respondent Sex and Item Order. ..115
15. Multivariate Analysis of Variance for ATL, ATG, AFI-L, AFI-G
by Item Order and Sex. ..........................1 1 6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
The primary purpose of this study was to examine whether heterosexuals tend to
hold different attitudes toward lesbians as a separate group and toward gay men as a
separate group, if the order that questions about lesbians and questions about gay men
were asked influence attitudes and attitude functions and whether attitude functions
predict attitudes. This study also explored the relationship of demographic and social
group differences to attitudes toward lesbians and gay men and the psychological
motivations behind these attitudes. Data from a 2003 national online self-report survey
of 411 adults (202 men and 209 women) revealed that women tended to hold similar
attitudes toward gay men and lesbians, whereas men were more likely to hold different
attitudes according to the sex of the target. Women’s attitudes toward gay men and
lesbians were more favorable than men’s attitudes toward these two groups. Men’s
attitudes toward gay men were the least favorable. Heterosexuals close to no gay men,
those who attend religious services weekly or more and Born-again Christians held the
least favorable attitudes. There were gender differences for closeness to gay men. Men
tended to be close to fewer gay men than their female counterparts. Men endorsed an
increased defensive function, while women endorsed the experiential-schematic function
more than men. Both of these attitude functions (psychological motivations) predicted
attitudes toward gay men and lesbians. Heterosexual women were more likely than men
were to agree that lesbians and gay men should be able to legally wed, should be legally
protected from job prejudice and discrimination, should be able to adopt children and
should be able to serve openly in the military. Of these four public policy issues, serving
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
openly in the military had the highest percentage gain, 13%, since 2000. Attitudes and
attitude functions were not dependent on the order in which questions about gay men and
questions about lesbians were asked. Implications for the importance of assessing
attitude functions and attitudes toward gay men and attitudes toward lesbians as separate
groups were discussed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
CHAPTER I
Introduction
Background of the Problem
There is much debate about how attitudes are formed and changed, the
motivation for holding different attitudes and how change in attitude is correlated to
behavior change (Collins, Kiesler, & Miller, 1969). Even so, researchers must continue
to add knowledge to this complex domain. There is an urgency for social scientists to
advocate for the well-being of groups that have been traditionally and systematically
devalued by the greater society. The deleterious effects of negative attitudes toward gay
men and lesbians are numerous (e.g., hate crimes and discrimination) and provide a
strong argument for better understanding these attitudes and creating and implementing
large-scale attitude change strategies.
Luhrs, Crawford, and Goldberg (1991) suggested that as the traditional nuclear
family is no longer the statistical norm and violence, crime and other societal ills are
rampant and highly publicized, any individual or group considered to be outside the
traditional norm may be symbolically linked with the transition and flexing of societal
boundaries. Many may view out-groups as contributing to these threatening changes in
society. Additionally, as the world’s available natural resources are rapidly shrinking,
those in the majority may continue to resist the acceptance of out-groups as acceptance
might entitle these out-group members to the limited resources typically reserved for
majority group members in any given society. Therefore, to deal with anxiety created
by this intrapsychic conflict, many continue to use social distancing, violence,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
bolstering and/or maintenance of negative attitudes, strengthening of in-group ties and
attachment to social or political groups that reject the out-group.
Sexual prejudice toward lesbians and gay men is alive and well in America.
Saying “I am gay” in the wrong place at the wrong time can still get one fired, harassed
or killed.
• The FBI’s Uniform Crime Reports show that hate crimes based on sexual
orientation constitute the fourth highest category, following only race, ethnicity
and religion. This despite the fact that anti-gay violence is typically
underreported (Human Rights Campaign, 2001).
• Lesbian, gay, bisexual and transgender (lgbt) youth “face daily threats to their
mental and physical health - ranging from anti-gay taunts to beatings - in their
schools and communities. [...] 31 percent of gay youth were threatened or
injured at school in just the last year.” Teenagers “hear anti-gay slurs such as
‘homo,’ ‘faggot’ and ‘sissy’ about 26 times a day or once every 14 minutes”
(National Mental Health Association, 2002).
• Fully 75% and 98% of lesbian, gay and bisexual college students reported verbal
insults directed at them, respectively (D’Augelli, 1989a; Herek, 1993). As many
as two million gay and lesbian teens are subjected to harassment and bullying.
School officials often refuse to address and sometimes actively encourage the
abuse (Human Rights Watch, 2001).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
• 40% of lgbt workers face hostility or harassment on the job. One in ten lgbt
adults has been “fired or dismissed unfairly from a previous job or pressured to
quit a job because of their sexual orientation” (Out & Equal, 2002).
While the recent slayings of lgbt people such as Matthew Shepard (Laramie,
Wyoming in 1998), Fred Martinez, Jr. (Cortez, Colorado in 2002), Gwen Araujo
(Newark, California in 2002), Sakia Gunn, (Newark, New Jersey in 2003) and Selena
Alvarez Hernandez (Council Bluffs, Iowa in 2003) have attained a high profile in the
media, it is important to remember that they are examples of a wider pattern of violence
against lgbt people.
The Henry J. Kaiser Family Foundation (2001) reported similar findings. Three
quarters (74%) of lesbians, gay men and bisexuals (lgb) reported having been the target
of verbal abuse because of their sexual orientation. Approximately one third (32%)
reported they have been the target of physical violence. Nearly half (41%) of lgb
respondents stated there is more violence directed towards lgb people in America today
compared to a few years ago and 40% said the level of violence has stayed the same.
Ninety percent (90%) of those surveyed, believe the government is doing too little to
protect them from discrimination. In a second survey conducted by The Henry J. Kaiser
Family Foundation (2001), 40% of heterosexual respondents believe the government is
doing too little to protect lgb people from discrimination and 24% believe the
government is doing too much. Herek (2002) cites a 1965 Harris poll where fully 75%
of the heterosexual respondents believed that homosexuals were more harmful than
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
helpful to American life. There appears to be evidence suggesting the treatment of
lesbians and gay men has not improved much over the past 20 years.
Blumenfeid (1992) makes the observation that homophobia or sexual prejudice
is detrimental to everyone, regardless of a person’s race, sexual orientation or religion.
Herek (2000) defined homophobia or his newly coined terminology, sexual prejudice as
negative attitudes toward an individual because of his or her sexual orientation. Sexual
prejudice: (a) confines individuals to prescribed gender roles, hindering individual
growth, creativity and authentic self-expression, (b) creates divisiveness and
maltreatment of lesbians and gay men, (c) inhibits people of the same sex forming close
relationships and bonds, especially between heterosexual men, (d) many times restricts
the interaction of lesbian and gay people with family members and co-workers, (e)
prevents heterosexuals from accepting and celebrating contributions of lesbian and gay
men in all domains of life and (f) thwarts appreciation of all forms of human diversity.
Altogether, some of these same effects can be seen as a result of racism, sexism
and the restriction of religious freedom. This study provides perspectives on the
descriptive questions of what are heterosexuals’ attitudes toward gay men and lesbians.
It will also provide new information for the functional question of why Americans hold
those attitudes. Knowing more about what use people have for their attitudes toward
gay men and toward lesbians will enrich our current understanding of how attitudes
form and why they change.
Further, more research is needed to examine the demographic and social
differences in attitudes and attitude functions; determine the relationship between
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
attitude functions and attitudes toward gay men and lesbians; explore gender differences
in attitudes toward gay men and attitudes toward lesbians; predict attitudes toward gay
men and lesbians from attitude functions; determine attitude stability by gender and
context effects; and determine attitude function stability by gender and context effects.
The results of this study may provide insights into strategies to improve attitudes toward
gay men and lesbians and other groups who are summarily discriminated against (e.g.,
racial, ethnic and socio-economic groups).
The primary goal of this study was to examine the relationship between
demographic and social differences in attitudes toward gay men, attitudes toward
lesbians and the functions these attitudes serve. The secondary goal was to examine the
relationship between gender and the order attitudes were assessed toward gay men and
toward lesbians, and attitudes and attitude functions.
This study is grounded in the functional (Katz, 1960) and neo-functional (Herek,
1987) approaches; gender/sexual identity and minority group paradigm models
proposed by Herek (2000); context-dependent attitude theory (Tourangeau and
Rasinski, 1988); theory on media effects on worldview and value system (Gross, 1991);
and the contact hypothesis (Allport, 1954).
Below is a review of the literature examining definitions of attitude, followed by
attitude change theory. The chapter concludes with attitude function theory in order to
lay the groundwork for new areas of inquiry.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
Definitions of Attitude
Asch (1952) commented on the state of group impression formation and change:
We know little today of the question at issue, mainly because of our failure to
study directly the process of impression forming. Therefore we are not in a
position to answer certain first questions such as: What are the organizational
properties of group impressions? In what respects do they differ among
individuals? What conditions determine their rigidity and lability? (p. 235)
The concept of attitude has played a key role throughout the history of American
social psychology. Pre-World War II efforts were targeted toward attitude
measurement and scaling. Postwar efforts, on the other hand, have focused on theories
about attitude formation and attitude change.
There are a myriad of definitions for the construct of attitude. For example,
Collins, Kiesier, and Miller (1969) defined attitude as “a process of individual
consciousness which determines real or possible activities of the individual in the social
world” (p. 6). Krech and Crutchfield (1948) defined attitude as “ ...an enduring
organization of motivational, emotional, perceptual, and cognitive processes with
respect to some aspect of the individual’s world” (p. 152). Eagly and Chaiken (1993)
defined attitude as a favorable or unfavorable evaluation of an attitude object (e.g.,
social group). From the exemplar perspective, Smith (1992) defined attitude as a
temporary mental representation that differs from one time to the next as different
subsets of exemplars become activated. Triandis (1971) stated an attitude is comprised
of three components: (a) the cognitive, or an idea used in thinking, (b) the affective, or
the emotion or feeling accompanying the idea, and (c) the behavioral, or disposition to
act based on the idea. Reardon (1981) claimed attitudes are (a) learned, (b) enduring, (c)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
affective evaluations and (d) related to human behavior; however, a growing body of
research calls into question the endurance of attitudes.
Theoretical discussions about socially significant problems have consistently
emphasized the behavioral relevance of attitudes. The recurring, unifying idea found
within each definition is the concern with the importance of individual differences or
the tendency for some individuals to behave differently from others under the same
circumstances. For the purposes of this study, attitude refers to a psychological
tendency that is expressed by a positive or negative evaluation of a particular entity or
attitude object (Eagly & Chaiken, 1993).
Attitude Representation and Change Theory
There are three main theories on how stereotypes and attitudes are represented in
memory. Some believe that groups are represented in memory abstractly with the group
being associated with traits or other identifying characteristics (Srull & Wyer, 1989).
Another school states that groups are represented purely in the form of specific
exemplars. The third and most popular view holds that both types of representations are
present—the mixed-representation view.
Research has been done to determine the conditions under which some people
access and use each type of representation (Fazio, Jackson, Dunton, & Williams, 1995).
Hastie, Schroeder, and Weber (1990) found that when a target person is a member of
more than one group, stereotypes from both groups are sometimes combined to create
new attributes. Researchers are beginning to conclude that impressions appear to be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
based on multiple sources of information instead of solely group or individual
characteristics (Gastardo-Conaco, 1990).
Influence of context on attitudes. Conceptualizing attitudes as interdependent
mental representations underlies two recent theoretical advances in the attitude domain.
The first, conceptualizes attitudes as mental representations activated in response to
being asked about attitudes or encouraged to act in response to one’s attitude. For
example, Tourangeau and Rasinski (1988) found that people asked about the
government responsibility toward helping the poor, later responded in favor of welfare
spending compared with responses of those who were prompted to think about the value
of individual effort.
The researchers concluded that if context plays a role in what attitude is
expressed, each individual may have many context-dependent attitudes and attitudes
could possibly be constructed in an ad hoc context-dependent manner. This implies that
the commonly held belief that attitudes are relatively long lasting and stable structures
requires further examination. Possibly, a person’s attitude toward a social object may
not be overwhelmingly favorable or unfavorable, but may vary in degree based on what
particular mental representation or episode is activated and retrieved.
The second recent development in the attitude domain is conceptualizing
attitudes as part of a mental representation that requires certain conditions to activate it,
especially when activation is automatic (Bargh, Chaiken, Raymond, & Hymes, 1996).
This model, which is also consistent with learning theory, posits that a positive or
negative attitude toward an object is linked to a representation of the object. Further,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
repeated activation of this associative link can strengthen the attitude making activation
of this particular evaluation automatic even at the mention of the attitude object. For
example, Macrae, Bodenhausen, Milne, and Jetten (1994) found that efforts to suppress
unwanted stereotypic thoughts led to increased use of those stereotypes later on. In this
case, trying to suppress activation actually led to activation.
In contrast to these findings, Kawakami, Moll, Hermsen, Dovidio, and Russin
(2000) examined whether sufficient training in negating stereotype associations of
skinheads, elderly people and different races would influence individuals to bypass the
activation of stereotypes following presentation of the stereotype category, thus
breaking the application of stereotypes for that category. They outlined three primary
conditions under which the application of stereotypes has been deterred: (a) when
individuals are motivated to be nonprejudiced (Monteith, Sherman, & Devine, 1998),
(b) when individuals experience a heightened awareness of egalitarian norms and
standards (Macrae, Bodenhausen, & Milne, 1997) and (c) when individuals have goals
that require them to gather unique information about an out-group member (Fiske &
Neuberg, 1990). Therefore, there is evidence to suggest that with the sufficient amount
of motivation and cognitive resources, individuals may not apply stereotypes.
Kawakami, Moll, Hermsen, Dovidio, and Russin (2000) hypothesized that practice,
which has been shown to play a significant role in automaticity, may also play a critical
role in reduction of stereotyping. They suggested that since low-prejudice people are
motivated to avoid the activation and application of stereotypes, they may have learned
through practice to control stereotyping.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
Using three studies, Kawakami, Moll, Hermsen, Dovidio, and Russin (2000)
found that subjects trained to say “no” to stereotyping broke automatic stereotype
activation. On the other hand, they found that subjects who received no training in
negating stereotypic associations demonstrated automatic stereotype activation. These
results were still apparent 24 hours after the training had ended. They concluded while
stereotype activation may be automatic for many individuals, it is not necessarily
inevitable or uncontrollable. Stereotype activation can be reduced.
They suggested these results are also in accordance with the ‘ thought- stopping ’
techniques used in cognitive therapy for emotional disorders, especially for the
treatment of depression (Beck, 1976). Kawakami, Moll, Hermsen, Dovidio, and Russin
(2000) suggested that future research should focus on determining the process by which
effective stereotype reduction occurs. Are the effects primarily due to motivational
factors with the goal being to not stereotype? Does stereotype reduction occur from the
cognitive effects involving the strengthening and weakening of category-trait
associations? Or, is it a combination of both motivational and cognitive influences?
This model presents some interesting and pragmatic research questions. Under
what conditions can activation of an attitude be blocked? Are there interventions that
can make a more favorable attitude toward a minority group become automatic? How
are attitudes really linked to behaviors? Do attitudes determine behavior or, under
certain conditions, do behaviors determine attitudes?
It should follow that when a person thinks about his or her attitude toward a
particular group and they encounter someone from that group who has characteristics
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
similar to how they view the group, behavior should match the activated attitude. On
the other hand, when the encountered group member is inconsistent with a person's
evaluation of the group, it is presumed that attitudes will not predict behavior.
Self-concept theory and attitudes. The study of the self is influenced by the
paradox that a person has stable enduring personality traits, yet he or she can act in
contrast to these traits in different social situations. Earlier research focused on the
stability of these traits, whereas current research has been examining the flexible,
context-dependent and fluid nature of the self (Linville & Carlston, 1994). These
formulations are consistent with recent attitude theory and research—a person’s self-
concept, as well as his or her attitudes toward others, depends on the context of the
interaction.
Social identity theory states that people seek positive aspects of their own group
in comparison to other groups (Tajfel, 1978). Lemyre and Smith (1985) suggested that
intergroup discrimination and negative attitudes toward other groups is a factor in
enhancing self-esteem. Schaller (1992) found that in order to maintain positive group
identity, prejudice and simplistic reasoning to defend negative attitudes are more likely
to occur if the outcome will result in their own group looking good.
Self-concept theory and research is important to the study of attitudes for
practical reasons. For example, Tesser’s (1988) self-evaluation model (SEM) focuses
on ways in which group members seek positive differentiation through religion, culture,
values, customs, art, language and other relevant dimensions. Under the SEM, if
Whites, as an in-group, value competition, ambition, intelligence and hard work, these
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
positive dimensions may be used to compare themselves with out-groups. Native
American culture does not necessarily value competition the same way Whites do. This
difference in values may contribute to holding a negative attitude toward Native
Americans by Whites. Paradoxically, Asian-Americans and Jews who generally value
some of the aforementioned White attributes, may be disliked and discriminated against
for possessing these socially valued attributes because they may threaten a White
person’s self-concept.
Attitude change theory. Yeb (1997) stated that people are not bom with
negative attitudes toward gay men and lesbians. These negative attitudes are formed
and acquired through interaction with teachers, family members, peers, friends, clergy,
etc. Negative attitudes are learned and maintained through communication. It is also
through persuasive communication that such attitudes can be altered.
Chaiken’s (1987) Heuristic Systematic Model (HSM) proposed that attitude
change can occur when: (1) favorable thoughts are generated about the message when it
is processed and/or (2) heuristics are used when processing the message (e.g., “this
expert is espousing positive attitudes about Latinos and experts can be trusted”).
Similarly, the Elaboration Likelihood Model (ELM) (Petty & Cacioppo, 1986) is based
on the theory that receivers of persuasive communication process the messages based
upon the relevance of the attitude object. There are seven principles related to the
process of attitude change in ELM.
First, people are motivated to hold accurate attitudes. Second, the situation or
context of communication (e.g., distraction) and internal state of the person (e.g.,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
motivation, mood, receiver involvement) influence the amount of message elaboration
and attitude change by the receiver. Third, the direction and degree of attitude change
is influenced by (a) persuasiveness of the argument (strong vs. weak), (b) amount and
type of elaboration by the receiver and (c) peripheral cues (attractiveness, likeability
and credibility of messenger) that tell the receiver what the response ought to be.
Fourth, ability and/or motivation of the receiver influence whether or not he or she
scrutinizes the quality of the message. Motivational factors include (a) receiver
involvement (people tend to elaborate on messages if the topic is important to them), (b)
need for cognition (the receiver’s tendency to think about particular topics) and (c) the
presence of multiple messengers communicating multiple arguments for the message
claim. Ability factors include (a) distraction and (b) prior knowledge of the topic.
Fifth, the ability or inability to process messages objectively influences attitude change.
Sixth, if a receiver is not motivated to or not able to analyze message quality, he or she
is likely to be persuaded on the basis of peripheral cues. On the other hand, if a receiver
is able to and motivated to examine message quality, he or she is likely to be influenced
by the quality of the message. Seventh, processing through the central route produces
longer lasting attitude change and greater behavior change.
According to HSM and ELM, it is the favorability of the thoughts generated by
the arguments in a persuasive message that bring about attitude change. The number
and positivity/negativity of thoughts evoked by a message depend on the extent to
which the person is willing or able to process the message. Individuals who are
motivated and able to elaborate on a message should have mostly favorable thoughts
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
when exposed to strong arguments and predominantly unfavorable thoughts when
exposed to weak arguments.
Both models predict that individuals who are able to pay attention to the
message and are motivated to process the message will show greater attitude change
compared with those who are less motivated and/or less able to pay attention to the
message. Both ELM and HSM assume that individuals who process a message
systematically are motivated to hold accurate or valid attitudes.
Using the tenets of ELM to produce attitude change toward gay men and
lesbians, the peripheral route to change would require the use of credible and likeable
speakers conveying the messages. The use of strong and persuasive arguments that
include the fact that the receiver is likely to either have a family member or close
associate or friend who is gay or lesbian, increasing the relevance of the topic for them
would be especially important.
Over the past 20 years, persuasion researchers such as Smith (1994), Chaiken
(1980) and Petty and Cacioppo (1981) have developed the dual-process model of
attitude change. One assumption of this model is that generally, people process
incoming information to a minimal extent and without much conscious effort.
Persuasive messages, if registered at all, are generally evaluated in terms of general
characteristics, such as message length, which may play a role in influencing attitudes.
On the other hand, high motivation and focused attention, characteristics held by people
processing something personally relevant and important, contribute to carefully
studying a persuasive message and systematically thinking about the implications of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
their attitudes. Cacioppo, Petty, Kao, and Rodriguez (1986) found that studying and
systematically processing a message leads to longer attitude change, more resilient
attitudes and a stronger correlation between attitude and behavior.
Tensor product model (TPM). Kashima, Woolcock, and Kashima (2000)
proposed a new model of group impression formation and change. TPM challenges the
commonly held belief in social science literature that the formation and change of
impressions are two separate processes. In other words, impressions are formed and
become stable entities and that changing impressions involves a different process.
Kashima, Woolcock, and Kashima (2000) suggested that the process underlying both
the formation and change of group impressions could be a single, learning process
described by the TPM. The model emphasizes the structured and contextual nature of
group impressions and the dynamic evolution of group impressions over time.
TPM assumes that through direct interaction with members of a social group or
indirect hearsay in interpersonal discourse or through media, perceivers form their
impressions about the group. The model also assumes that the context (e.g., social
situation, information source, affective state of the self) in which information about a
group was obtained is important in impression formation and change. The information
about the group is analyzed and encoded into specific features (e.g., skin color, friendly
personality). The central assumption of TPM is that each new experience or
representation with a group or group member is superimposed on preexisting
representations. With the passing of time, the memory of specific representations is
forgotten. TPM assumes that group impressions are dynamic cognitive configurations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
Further, group impressions represent the social perceivers’ flexibly structured and
constantly evolving understandings about social groups.
Using the TPM, Kashima, Woolcock, and Kashima (2000) found that stereotype
change was mediated by the extent to which a stereotype-inconsistent group member is
individuated. TPM implies that an individuated group member’s inconsistent
information does not influence the stereotype as much as a nonindividuated member’s
equally inconsistent information. They also found that stereotypes changed more when
inconsistent information was dispersed across a great number of group members than
when concentrated in a few individuals. This is likely true because extremely atypical
individuals are probably individuated or sub typed, having little or no effect on
stereotype change because they are cognitively isolated, leading to the insulation of the
stereotype from change.
Influence of disconfirming information about stereotypes on stereotype change.
Hewstone, Hassebrauck, Wirth, and Waenke (2000) examined how slightly
disconfirming stereotype information dispersed across several group members versus
strongly disconfirming stereotype information dispersed across a few group members
influences stereotype change.
Hewstone, Hassebrauck, Wirth, and Waenke (2000) found that when subjects
focused on similarities (versus differences versus a control) between group members,
and after they were exposed to stereotype disconfirming information spread across
many group members, the greatest reduction in stereotyping occurred. They concluded
that stereotype change might be strongly impacted by disconfirming group members
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
who are also seen as typical of a group because the disconfirming information is spread
amongst many.
Many of the newer models that take into consideration the interactive influence
of multiple processing schemas, episodes, situations and norms replace the old
assumption that a single attitude, stereotype or prejudice will be activated and used in
every situation. Attitudes often fluctuate according to a variety of sometimes
conflicting factors (Smith, 1992). These models and findings also suggest that
impression and attitude formation and change are very complex constructs that require
further controlled and practical research.
Summary o f attitude formation and change theory. There are three main theories
on how stereotypes and attitudes are represented in memory. Some believe that groups
are represented in memory abstractly with the group being associated with traits or
other identifying characteristics (Srull & Wyer, 1989). Another school states that
groups are represented purely in the form of specific exemplars. The third and most
popular view holds that both types of representations are present—the mixed-
representation view.
There is some evidence suggesting that each individual may have context-
dependent attitudes. Therefore, attitudes toward an attitude object may not be
predominantly favorable or unfavorable. Instead, they may vary based on what
particular mental representation or episode is activated or retrieved. Earlier research
focused on the stability of personality traits, attitudes and behaviors. Current research
has been examining the flexible, context-dependent and fluid nature of the self and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
attitudes. There is a growing body of evidence suggesting that a person’s attitude as
well as his or her behavior toward others depends on the context of the interaction.
Kashima, Woolcock, and Kashima (2000) proposed a new model of group
impression formation and change. They suggest that the process underlying both the
formation and change of group impressions could be a single, learning process. The
model emphasizes the structured and contextual nature of group impressions and the
dynamic evolution of group impressions over time. This innovative model requires
empirical testing to validate its tenets.
Other research has identified many factors contributing to attitude
change—motivation, mood, ability to elaborate on the message, personal importance of
attitude object, strong versus weak persuasive arguments and the credibility and
likeability of the messenger.
Attitude Function Theory
Why people are more prejudiced toward one group than toward another can be
explained by the functional approach to attitudes (Katz, 1960; Herek, 1986a).
According to the functional approach, people hold certain attitudes because they receive
some psychological benefit from doing so. Two people can have the same attitude that
serves different functions for each person. For example, one person might support the
ban against gay men and lesbians serving in the military because he or she believes that
homosexuality is a sin. Whereas, another individual may hold the same attitude
because men in his fraternity support the military ban and the individual chooses to
support this attitude in order to fit in with a specific social group. An attitude can elicit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
a single function or multiple ones. Additionally, attitudes about different groups or
objects may serve different functions. By understanding the specific function or
functions an attitude serves for an individual, attempts can be made to change the
attitude.
During the 1950s and 1960s, a variety of attitude functions were proposed by
social psychologists. Although the labels ascribed to functions might have been
different, after thorough examination, there was much overlap in the definitions of
functions. Katz (1960) proposed four attitude functions, while Smith, Bruner, and
White (1956) proposed three attitude functions. The utilitarian function (Katz, 1960)
states that an attitude is held to minimize punishments and maximize rewards from the
person’s environment. The second or knowledge function (Katz, 1960) states that an
attitude is held in order to help make sense of one’s world to provide meaning and
consistency. Similarly, Smith, Bruner, and White (1956) proposed the object appraisal
function, which states that we organize the objects in the world to satisfy our interests
and ongoing concerns. The ego defense function, described by Katz (1960) and
similarly the extemalization function described by Smith, Bruner, and White (1956) are
held as coping strategies for intra-psychic conflict by denying, repressing or projecting
his or her fears onto others in defense of one’s self-concept. The fourth function
proposed by Katz (1960), the value-expressive function allows the expression of values
important to one’s self-concept. The third function proposed by Smith, Bruner, and
White (1956), the social adjustive function, maintains that an attitude might be held to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
facilitate interpersonal relationships by gaining social acceptance or aligning oneself
with groups they have or want membership in.
Until the mid-1980s, relatively little research was conducted on attitude
functions. Maio and Olson (1994) suggested that the scarcity of research was due to
difficulties in operationalizing attitude functions. Overlap of concepts and definitions
may also have contributed to the lack of empirical research. Additionally, psychology
during the 1960s and 1970s turned its focus on situationist and interactionist
explanations of personality and the functional approach was identified as a personality
trait approach, treating attitudes as dispositions. For example, empirical inquiry usually
focused on personality variables as the primary factors in determining attitude
functions. It was assumed, for example, that a defensive person would have attitudes
serving a defensive function, regardless of the attitude object. Functionalism was
associated with psychodynamic theories of personality, which many social
psychologists have rejected. The functional approach to attitudes never developed into
a theory that specified that attitude functions could be assessed, nor did it identify under
what conditions each function would prevail. Most importantly, for over 30 years, the
functional approach was never empirically validated due to the absence of valid and
reliable methods for assessing functions.
Development o f the attitudes function inventory. Over the past 20 years, there
has been a renewed interest in the functional approach to attitudes. First, Herek (1987)
created the Attitudes Function Inventory (AFI). The AFI, a 10-item questionnaire, was
developed and used to assess the functions served by attitudes toward gay men and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
lesbians as well as toward people with AIDS, mental illness and cancer. Using a
content-analysis procedure, the presence of four factors or functions was indicated:
experiential-schematic or knowledge, defensive, social-expressive and value-
expressive.
First, Herek (1987) examined attitudes toward gay men and lesbians because at
the time of the study, this target group was relatively visible on the host college campus.
Herek (1987) surmised that the topic of homosexuality would elicit strong attitudes,
both hostile and favorable. Additionally, he assumed that attitudes toward gay men and
lesbians would be deeply felt and would serve a wide range of functions (Herek,
1984a). In a second experiment, Herek (1987) examined attitudes toward people with
AIDS, mental illness and cancer to further validate his initial findings.
Herek (1987) found that there are four general categories of attitude functions
that are similar to the functions proposed in the 1950s and 1960s. He concluded that
holding favorable or unfavorable attitudes toward these four groups can serve the value-
expressive, social-expressive, defensive and/or experiential-schematic functions.
The value-expressive function serves a need for personal identity. The value-
expressive attitude provides a way of expressing the value and theory affirming whom
one is in terms of good and bad, right and wrong. An example of a value-expressive
attitude is “I have negative attitudes because the Bible clearly states that homosexuality
is a sin.” Herek (1986b) purported that changing a value-expressive attitude may
happen by hearing a respected authority, in this case a minister or Biblical scholar, who
offers an interpretation of the Bible that does not include condemnation of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
homosexuality. As religious affiliation is a standard demographic variable found in
virtually all psychological and other social science research, testing this hypothesis
would be relatively straightforward.
The social-expressive function defines group boundaries with the self on the
inside or as part of the majority and the gay man or lesbian on the outside. To fit and
conform, the person adopts a negative attitude toward the minority. Rogers (1998)
contended that there is a likely interaction between the value-expressive and social-
expressive function, as can be observed by radical Christian extremist groups that rely
on the creation of an “other” (e.g., homosexuals) who they contend are wrong and bad.
Herek (1986b) suggested that altering a person’s idea of social norms could
change someone’s negative social-expressive attitude. For example, an openly gay man
or lesbian in a social group can help to change the very nature of the social norm within
that group.
The third type, the defensive function, is rooted in anxiety created by conflicting
feelings about accepted gender roles and/or sexual orientation and the person’s own
contrary experience. In other words, who the person should be by majority standards is
not who they are. For example, a person may have same-sex erotic desires or realize he
or she is acting atypical to prescribed gender roles, and he or she cannot accept these
conflicts. To deal with this conflict, the person may hate or fear the external object.
Observable manifestations of defensive attitudes might come in the form of
hostility expressed through verbal or physical assaults on the attitude object. Herek
(1984a) suggested that many of the correlates found to be related to negative attitudes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
toward gay men and lesbians—authoritarianism, cognitive rigidity, intolerance of
ambiguity and dogmatism—are also personality traits that generally indicate higher
levels of defensiveness. Herek (1986a) also concluded that the defensive function helps
explain the finding that there are greater negative attitudes toward gay men and lesbians
of the same sex as the respondent because they pose a greater threat to one’s homoerotic
conflicts. Further, Cuenot and Fugita (1982) found that in the presence of a gay man or
lesbian, subjects’ speech tended to become more rapid and they espoused more sexual
conservatism in comparison to when a heterosexual was present during the interview.
When the perceived homosexual was of the same sex as the subject, their avoidance
reaction was greater than in comparison with an opposite sex homosexual. Due to the
deep unconscious nature of this attitude type, Herek (1986b) suggested this is the most
difficult to influence. His proposed strategy for change includes teaching new
meanings for masculinity and femininity.
The fourth type, the experiential-schematic or knowledge function, helps us to
make sense of the world by categorizing things according to our past experiences.
Opinion poll data consistently shows that people who indicate they know at least one
openly gay man or lesbian are more likely than others to hold more positive attitudes
toward gay men and lesbians. Herek (1986b) drew on racial and ethnic attitude change
research and concluded that in order for negative attitudes toward gay men and lesbians
to improve, the type of contact is important: (a) the experience with a gay man or
lesbian needs to be ongoing and intimate, (b) the relationship between the lesbian and
gay person and heterosexual person must be cemented by similar beliefs and values and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
(c) the relationship must be based on shared goals and cooperation, rather than
competition. Situations that typically fit these criteria occur between close friends, co
workers and family members. This also requires that gay men and lesbians are open
about their sexual orientation. Lack of basic civil rights for gay men and lesbians as
well as the fear of rejection, harassment and violence still discourages many gay men
and lesbians to come out.
Shavitt (1989) suggested that the construct validity of the AFI might be
restricted to the attitude objects that Herek examined. To evaluate Shavitt5 s suggestion,
Anderson and Kristiansen (1990) examined the factor structure of subjects’ AFI
responses on a wide variety of attitude objects (Toyota cars, air conditioners, abortion
and gay rights) that would likely elicit a variety of attitude functions. From their
findings, Anderson and Kristiansen (1990) concluded the AFI could be used to measure
the function of people’s attitudes, regardless of the specific attitude object.
Person, domain and situational characteristics. Herek (1987) noted the
importance of understanding functions as a reflection of the interaction of
characteristics specific to individuals, domains and the situations in which attitudes are
formed and expressed. First, there are many personality correlates of attitude functions.
For example, attitudes serving a defensive function may be more likely found among
individuals who exhibit intrapsychic conflict associated with the attitude domain (e.g.,
hyper conformity to gender role expectations in the case of attitudes toward gay men
and lesbians) as well as externalizing defenses to deal with those conflicts. The
dominant personality characteristic for those who primarily hold a social-expressive
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
attitude is a heightened awareness of others5 response to their actions and beliefs and
the need for that reaction to be favorable. In contrast, those holding a value-expressive
attitude may have less need for acceptance by one’s own social group. Being true to the
values he or she holds toward the reference group is of utmost concern to these
individuals. Those holding experiential-schematic attitudes tend to combine an ongoing
awareness of their own internal states while integrating sensitivity to those of the
reference group.
Second, attitude functions are also influenced by the socially constructed
meanings of the attitude object. Since homosexuality has been defined in moral, legal
and ethical terms it can easily elicit the value-expressive function. Formal (churches
and political organizations) and informal (social groups) groups have taken a stance
toward gay men and lesbians; this facet of the domain may elicit a social-expressive
attitude function. Further, homosexuality has been closely linked to cultural constructs
of gender. So, attitudes toward gay men and lesbians can serve a defensive function.
The experiential-schematic function is activated when gay men and lesbians come out
to their friends, family and colleagues.
The third factor to consider when analyzing attitude functions is the setting in
which attitudes are formed and expressed. The time, place and social context may
influence the attitude function expressed. For example, a man who would usually
endorse a social-expressive function may score highly on the defensive function
because he was just told he has homosexual tendencies and he becomes anxious and
angry. Considering the interaction of all three characteristics—personality traits, social
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
construction of a minority group and the situation—is important to the study of attitude
functions.
Summary of attitude function theory. There is consistent evidence illustrating
there are four primary motivations behind holding attitudes toward different attitude
objects. The four functions are (a) defensive, (b) experiential-schematic, (c) value-
expressive and (d) social-expressive.
Overview o f the Study
As discussed at the beginning of this chapter, the primary purpose of this study
was to examine whether heterosexuals tend to hold different attitudes toward lesbians as
a separate group and toward gay men as a separate group. Examining the relationship
of demographic and social group differences (e.g., # of close relationships to gay men
and # of close relationships to lesbians as separate independent variables) to attitudes
toward lesbians and gay men as well as to attitude functions was also a primary goal of
this study. Another important goal was to determine if attitude functions predict
attitudes. A secondary goal was to examine if the order that questions about lesbians
and questions about gay men were asked influence attitudes and attitude functions.
Assessing overall agreement with specific public policy issues for gay men and lesbians
and whether or not there would be gender group differences in opinions was also a
secondary goal of this study.
This study was intended to further the functional (Katz, 1960) and neo-
functional (Herek, 1987) approaches; gender/sexual identity and minority group
paradigm models proposed by Herek (2000); context-dependent attitude theory
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
(Tourangeau & Rasinski, 1988); theory on media effects on worldview and value
system (Gross, 1991); and the contact hypothesis (Allport, 1954).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
CHAPTER II
Review of the Literature
This chapter begins by reviewing the literature on demographic and social
differences on attitudes toward gay men and lesbians. The next section examines the
contact hypothesis and how close relationships with gay men and lesbians tend to
influence heterosexuals’ attitudes toward gay men and lesbians. This section is
followed by literature on the effects of the media on attitudes. Next, attitude function
literature is reviewed to lay the groundwork for the importance of determining if
attitude functions predict attitudes. The next section includes studies examining gender
differences and differences in item order on attitudes toward gay men and lesbians. The
chapter concludes with a synthesis of the review of the literature, purpose of the study
and research questions and hypotheses.
Review of the Literature on Attitudes Toward Gay Men and Lesbians
Kuklinski, Riggle, Ottati, Schwarz, and Wyer (1991) stated that more than three
decades of research into the professed democratic values of American citizens has
consistently shown that there is high endorsement of principles in the abstract, and
considerably less support for specific applications when it comes to the fair treatment of
gay men and lesbians. To illustrate this point, Rogers (1998) pointed out that in 1977, a
majority (56%) of the public was against job discrimination toward gay men and
lesbians, and in 1996, this figure rose to 84%. However, in 1996, there was still less
(66%) support for specific laws against job discrimination toward gay men and lesbians.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
In 2001, that number rose to 76% (The Henry J. Kaiser Family Foundation). Further,
Rogers (1998) cited a 1977 Louis Harris & Associates poll that found 48% of
Americans asked thought that gay men and lesbians should not be allowed to become
psychiatrists and social workers. In the same poll, only 12% believed that gay men and
lesbians should not be sales people. It appears that occupations with low personal
contact receive less public opposition for gay men or lesbians to be in those positions.
Rudolph (1990) found that within a group of counselors and counselor trainees, there
was support for gay and lesbian civil rights, but there was a lower acceptance for
homosexuals in sensitive professional positions, like teaching. This finding may be due
to the perceived lack of moral character that many stigmatized groups experience.
Further, Rogers (1998) cited another poll from 1992 that found only 73% of
those Americans polled felt that gay men and lesbians should have the right to stay in
any hotel or motel they could afford. With support from hundreds of polls from 1965
through 1996, Rogers (1998) concluded that equal rights for lesbians and gay men are
not very important to the American public.
Sex, age, education level, religiosity, past interactions with gay men and
lesbians, adherence to traditional ideologies of family and gender, perception of
friends’ agreement with one’s own attitudes and residence. In a series of three studies
with college students at six different universities, Herek (1988) found that between 45%
and 62% of the variance in attitudes toward lesbians and gay men was explained by four
variables. Perceived agreement in attitudes toward homosexuality by friends, church
attendance and religious ideology, contact with lesbians and gay men and traditional
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
gender-role and family ideologies all played a significant role in attitudes toward gay
men and lesbians, but no single correlate of heterosexuals’ attitudes emerged as more
predictive of ATLG scores than did others.
He also found that perceiving oneself as not fitting stereotypes of either
masculinity or femininity was also associated with more favorable attitudes toward gay
men and lesbians. Further, perceptions of perceptions of dissimilarity between men in
general, women in general and gay men in general were significant predictors of
attitudes. However, there was not a significant difference between the perception of
respondents’ dissimilarity between lesbians and their attitudes toward gay men and
lesbians. Interestingly, psychological androgyny has not been consistently related to
attitudes toward gay men and lesbians.
Rogers (1998) found that women, younger respondents, those more educated,
those who do not claim a religion and those who live in New England and the Pacific
Coast had significantly more positive attitudes toward gay men and lesbians. The
Henry J. Kaiser Family Foundation (2001) found that younger adults (age 18-29) and
seniors (ages 65 and older) held divergent views on lgb moral and policy issues. For
example, the majority of younger adults supported gay and lesbian unions, marriage and
adoption rights (68%, 60% and 55% respectively), while far fewer seniors supported
those policies (27%, 25% and 25% respectively). Seniors (57%), men (57%) and those
with a high school education or less (57%) believed that homosexual behavior is
morally wrong. On the other hand, a smaller percentage of younger adults (44%),
women (48%) and those with at least some college education (46%) endorsed the same
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
statement. Approximately 60% of non-Evangelical Christians, 59% Catholics and 63%
of those with no religious affiliation thought lesbian, gay and bisexual couples could be
just as good parents as heterosexual couples. Evangelical Christians were divided
evenly on the issue: 47% agreed and 47% disagreed. Nearly three-quarters (74%) of
Evangelical Christians surveyed (The Henry J. Kaiser Family Foundation, 2001)
believed homosexual behavior is morally wrong, compared to 48% of non-Evangelical
Christians, 40% of Catholics and 20% of those with no religious affiliation. These
results suggest that it is critical for progressive religious leaders to dialogue with leaders
of traditional religions to discuss the negative influence their prohibition against
homosexuality has on gay men, lesbians and their families. Further, more research
needs to be done to determine if educated people tend to learn acceptance of gay men
and lesbians through their academic endeavors. Or, do those who accept gay men and
lesbians tend to pursue advanced education more than those who hold negative
attitudes? Further research assessing what types of people are more open to positively
changing their attitudes toward minority groups will guide the future work of
progressive individuals who wield political influence.
Using Herek’s (1988) Attitudes Toward Lesbians and Gay Men scale (ATLG)
and additional items, D ’Augelli (1989b) surveyed more than 100 residence assistant
(RA) applicants to learn more about the attitudes of students who hold leadership
positions and have a fair amount of contact with other students. He found that, overall,
heterosexual men had significantly more negative attitudes than did heterosexual
women. Men’s and women’s views of lesbians did not differ significantly. However,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
men’s attitudes toward gay men were significantly more negative than women’s
attitudes toward gay men. Less than 50% of those surveyed knew a gay men or lesbian
well. Those who reported knowing gay men or lesbians were less homophobic than
those who did not. Additionally, 100% of the RA candidates had heard disparaging
remarks about lesbians or gay men and 90% believed that harassment and threats of
violence or physical attacks against gay men and lesbians are likely. He concluded that
greater exposure to lesbians and gay men would probably decrease negative attitudes.
However, because gay men and lesbians on this college campus generally felt unsafe
due to their sexual orientation, many chose not to come out as gay or lesbian.
For many years, it was assumed that with age, people become more conservative
in their views. However, Wilson (1996) stated that many researchers have found that in
regard to racial and ethnic prejudice, younger adults are usually less prejudiced than
older ones and remain that way as they age. Using cohort analysis to examine the
contribution of age, period and cohort effects on attitudes toward race, Wilson (1996)
found that post-World War II cohorts (group shares significant life event and are bom
within 10 years of one another), were generally less prejudiced than pre-War ones. But,
he also found that the cohort bom after 1960 was actually more prejudiced toward
Latinos, Asian Americans and Jews, with no differences in regard to African Americans
than the pre-World War II cohorts. Consistent with those results, Ryder (1965)
suggested that new cohorts provide the opportunity for social change to occur, but they
do not cause the change. Stimson (1991) stated that cohort replacement, or the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
replacement of older birth cohorts by younger ones, appears to have an important
influence on public opinion, especially on social and cultural issues.
Rogers (1998) analyzed the proportions of change due to intra-cohort effects and
generational replacement in relation to attitudes toward civil rights, civil liberties and
public acceptance of gay men and lesbians. As expected, he found that cohort
replacement contributed significantly toward the change in public attitudes toward gay
men and lesbians. Interestingly, intra-cohort change played a more significant role in
changing attitudes in recent years than did cohort replacement. These findings suggest
that the process of intra-cohort attitude change toward gay men and lesbians in recent
years is a very important factor.
...the attitudes of new cohorts are dependent, to a large extent, upon their
common socialization into society. The normalizing pressures against
homosexuality are so strong for children and young adults that it is likely that
only a fairly small percentage of each cohort thinks independently enough to
break away from them. For broad change to occur, especially on measures that
have proven resistant to change, such as many civil rights and public acceptance
measures, it appears that there must be substantial numbers of people in the
population changing their minds (Rogers, 1998, p. 275).
Cognitive and affective bases for attitudes toward gay men and lesbians. Using
clinical psychology students and business students (32% were between 31 and 50 years
old) from Sweden, Emulf, Innala, and Whitam (1987) examined the cognitive and
affective bases of attitudes toward gay men and lesbians. To measure affect, an
adjective checklist of items describing homosexuality was administered. The stimulus
was a short story about a youthful romance, with the protagonists either being two
heterosexual lovers, lesbian lovers or two gay male lovers. In the study, the adjective
checklist always followed the story, but both were randomly placed before or after the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
attitude scale. Emulf, Innala, and Whitam (1987) found primarily two affective
factors—homophobic guilt and homophobic anger—accounted for 48% and 31% of the
variance explained, respectively. Of these two factors, only anger was significantly
correlated with attitude toward homosexuality (r=-.73) suggesting that those who
experienced anger in response to homosexuality also have negative attitudes toward
homosexuality. Overall, women had more positive attitudes than men, business
students had more negative cognitive attitudes, business students and male subjects who
read either homosexual story were significantly angrier than those who read the
heterosexual story.
Emulf, Innala, and Whitam (1987) suggested that homophobic anger manifests
itself in the form of violence against gay men and lesbians and contempt for their civil
rights and civil liberties. To explain why the guilt was not correlated with negative
attitudes, Emulf, Innala, and Whitam (1987) proposed an interesting interpretation.
They noted that in Sweden it is socially desirable to be publicly nonprejudiced. So,
Swedish people have been taught to learn positive attitudes toward gay men and
lesbians at a cognitive level, while they might repress negative emotional reactions.
These repressed negative reactions may lead to the feelings of guilt and shame. They
also speculated that those with both negative cognitions and negative emotions toward
homosexuals might also experience guilt and shame because they have repressed their
own homoerotic feelings, paralleling Herek’s (1984) defensive-expressive function of
attitudes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
Aguero, Bloch, and Byme (1984) assessed college students’ attitudes, behaviors
and experiences in relation to homosexuality. They found that the greatest dislike
toward homosexuals was in subjects who reported both negative affect and a belief that
homosexuality is a choice or learned behavior. Further, subjects who reported avoiding
social situations where homosexuals are present was correlated with both negative
affect and a belief that homosexuality is due to genetic factors.
Attitude strength and attitudes toward gay men and lesbians. Zuwerink and
Devine (1996) investigated how strong negative attitudes toward gay men and lesbians
are resistant to change and took an in-depth look at the resistance process. According to
Krosnick (1988), strong or important attitudes are more likely to guide behavior, to be
stable over time, and are more likely to influence information processing and social
judgments compared with less important attitudes. Zuwerink and Devine (1996) found
that strongly held negative attitudes toward gay men and lesbians in the military were
resistant to a pro-gay men and lesbians in the military argument. Participants with
strong negative views toward lesbian and gay men in the military responded to the
positive message by becoming angry and irritated. In addition, they expressed angry
and irritated thoughts about the speaker and his or her message.
Zuwerick and Devine (1996) concluded that participants who considered their
attitude high in personal importance, as opposed to less important, were more resistant
to attitude change. Their results also support the assumption that for attitudes that are
important to individuals, affective and cognitive processes moderate resistance to
persuasion. These findings support Festinger’s (1957) and many other theorists’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
assumption that attitude importance is associated with greater motivation to resist
change. That is, as attitude importance increases, attitude change probably becomes a
less likely dissonance reduction strategy. Instead, people may find other ways to reduce
dissonance, such as producing attitude bolstering thoughts, or even using aggression
toward the attitude object.
Ajzen (2001) reported the strength of an attitude varies over the life cycle with
the mid-life representing the strongest representation. Further, strong attitudes were
more resistant to change and were found to be more accessible than weaker attitudes.
These findings have implications for identifying target groups (people
undecided about or holding moderately negative attitudes) to focus on for attitude
improvement toward gay men and lesbians. Additionally, creating strategies to lower
the importance or strength of attitudes toward gay men and lesbians for those who hold
strong, negative attitudes may create opportunities to eventually influence those
attitudes.
Influence o f educational strategies on attitudes toward gay men and lesbians.
Christensen and Sorensen (1994) studied the effects of a program consisting of films,
readings, discussion and role-plays on the attitudes of child and youth worker students
toward gay men and lesbians. On measures of affect, cognition and behavior, the
experimental group showed more accepting views of gay men and lesbians than the
control group. Four weeks later, a second post-test was administered and the positive
gains in attitudes made initially had almost disappeared. This lack of durability in
attitude improvement may be due to the brevity (6 hours) of the training or to the reality
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
that attitudes about things that hold value to someone, are durable and hard to
permanently change.
Other investigators have developed and implemented similar interventions
aimed at increasing knowledge about gay men and lesbians. These interventions
include exposing heterosexual participants to gay men and lesbians through speaker
panels, film presentations depicting positive images of gay men and lesbians and
presentations including historical and political information about the plight of these
groups. Presentations about human sexuality, using didactic and experiential
techniques, have positively modified attitudes toward sexuality, including attitudes
toward gay men and lesbians (Kilman, Wanless, Sabalis, & Sullivan, 1981; Voss,
1980). The durability of these improved attitudes is in question because these studies
did not incorporate a long-term follow-up component to their research.
Dating back to the Stonewall Bar uprising in New York in 1969, where Igbt
clientele attacked police during a raid in response to years of oppression and police
brutality against lgbt people. In response to the perception that law enforcement in the
United States has historically been heterosexist, sexual orientation and diversity
trainings have become mandatory. However, most police trainings do not address the
individual differences and special needs of minorities and gay men and lesbians (ACA
and PERF, 1992). It is commonplace for police officers to view these trainings as
forms of punishment imposed upon them through outside political pressures instead of
opportunities to learn about different cultures and how personal biases may negatively
impact their ability to effectively do their jobs. Additionally, diversity or cultural
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
sensitivity trainings, especially about gay men and lesbians, are anathema to the police
stereotype—Hyper masculine, authoritarian, prejudiced and bigoted, needing to be in
control and cynical by nature.
In an assessment of seven California police agencies, Stewart (1997) found a
significant difference in heterosexual administrators’ perceptions of police culture and
that of gay and lesbian officers when it relates to gay and lesbian issues. For examples,
the majority of administrators claimed that homosexuality was “not an issue” and that
officers were judged solely on merit. On the other hand, gay and lesbian officers
expressed a perception that the climate was dangerous for openly gay and lesbian
officers.
Stewart (1997) also found the agencies that publicly recognized the
achievements of gay and lesbian officers through press releases or awards and
supported a culture where open discussion of partners of gay and lesbian officers was
the norm scored significantly lower on homophobia scales. Police recruits scored the
highest on homophobia scales. Openly gay or lesbian police personnel or heterosexual
police personnel who have a gay or lesbian child proved to be the most effective
instructors and panel members in terms of reducing homophobia.
From his results, Stewart (1997) concluded that the effectiveness of sensitivity
training is highly dependent upon the police culture. “To make a law enforcement
agency less heterosexist and more accepting of gays and lesbians it is important to bring
visibility to gay and lesbian officers and to validate their relationships” (Stewart, 1997,
p. 334). Everyday conversations need to include discussions about gay and lesbian
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
relationships and behavior. Law enforcement administrators and leaders need to model
attitudes and behavior that embrace sexual diversity and diversity in all its forms.
Further, sexual orientation trainings need to be relevant to police officers. For example,
each training should provide discussion time for specific police scenarios in which
sexual orientation had a potential impact on the outcome.
Influence of out-group versus in-group messenger on attitudes toward gay men
and lesbians. Clark and Maass (1988) conducted experiments to compare the in-group
minority (differing from the majority only in terms of their beliefs) and out-group
minority (differing from the majority in both their beliefs and their social
categorization) message delivery and its influence on attitudes toward gay men and
lesbians. Those in the in-group (majority beliefs and same social categorization) were
randomly assigned to four groups. Two of the in-groups heard a message in favor of
lesbian and gay rights from either a heterosexual in-group minority member or a
homosexual out-group minority member. The other in-groups heard a message arguing
against gay and lesbian rights by an in-group minority member or an out-group minority
member. Clark and Maass (1988) found that an in-group minority’s influence on
attitudes toward gay and lesbian rights is significantly greater than the influence that
out-group minorities exert, with either a pro- or anti-gay and lesbian rights stance. The
notion that out-group minorities are less influential in changing negative attitudes
toward gay men and lesbians has powerful implications for the gay and lesbian
movement. It suggests that heterosexual allies can more readily influence attitudes
toward gay men and lesbians than gay and lesbian activists themselves. This finding is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
important for professionals who are influencing the hearts and minds of Americans
everyday —psychologists, teachers, politicians, social advocacy organizations and
clergy. It is also important to conduct similar research with different races and sexes as
spokespeople.
Many of the conflicting results found in attitude research about gay men and
lesbians are similar to findings on attitudes toward race and sex. As attitude scales and
methods of attitude measurement improve, it will be important to continue researching
what individual, demographic and social differences in people contribute to improving
attitudes toward racial minorities, women, gay men, lesbians and other populations. It
is also important to continue examining why people change their attitudes.
Summary o f attitudes toward gay men and lesbians. There is a significant body
of research illustrating attitudes toward gay men and lesbians have been improving over
the past 50 years. More favorable attitudes have been generally associated with being
younger, female, having close relationships with gay people, having higher levels of
education and income, less religiosity, lack of adherence to traditional family and
gender roles and living on the west or northeast coasts primarily in urban areas. Most
studies have assessed attitudes toward gay men and lesbians as one reference group.
Additionally, research suggests that educational interventions may have short-term
favorable effects on attitudes toward gay men and lesbians. It also appears that
heterosexuals’ delivering a pro-gay or lesbian message are more effective at favorably
influencing the attitudes of their heterosexual counterparts. There is a need for future
research to determine if this attitude change is long-term or just temporary.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
Review o f the Literature on the Contact Hypothesis
Dovidio, Brigham, Johnson, and Gaertner (1996) pointed out that, traditionally,
researchers have assumed that stereotyping promotes prejudice, which, in turn,
promotes discrimination. In other words, beliefs about what a group is like are assumed
to determine the level of liking for the group, which then dictates favorable or
unfavorable action toward the group or individual group member. This assumption has
led to the widely endorsed contact hypothesis, which purports that if one can positively
alter beliefs about a certain group through education or a positive experience with an
individual in the group, then discrimination will decrease (Allport, 1954).
Millar and Millar (1996) and Millar and Tesser (1986) hypothesized that direct
experience with an attitude object would tend to produce more affective reactions than
would indirect experience with the object. Conversely, they hypothesized that indirect
experience with an attitude object would produce more cognitive reactions, than would
direct experience. Additionally, they predicted that attitudes produced after direct
experience with an attitude object would predict consumatory behavior (a behavior
performed only for the activity or process involved whereby the focus is on the intrinsic
enjoyment of the activity). They also predicted that instrumental behavior (a behavior
performed to accomplish a goal beyond the activity involved in performing It) would
follow indirect experience with an attitude object. Millar and Tesser (1986) found that
consumatory behavior is more affectively driven and instrumental behavior is more
cognitively driven.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Millar and Millar (1996) and Millar and Tesser’s (1986) studies are important
because if direct and indirect experience with an attitude object (one on one interactions
with an out-group member versus reading about an attitude object) is linked to the
degree to which an attitude is affectively or cognitively based, this could have
implications for attitude change strategies. Millar and Millar (1996) found that attitudes
formed through direct experience produced more affective reactions and made better
predictors of affectively driven behaviors (consumatory). Alternatively, attitudes
formed through indirect experience produced more cognitive reactions and made better
predictors of cognitively driven behaviors (instrumental). Millar and Millar (1996)
suggested that there are a number of differences between direct and indirect experience
that could account for direct experience producing more affective responses. One
difference is that direct experience with an attitude object allows for the opportunity to
receive feedback on how the object is impacting the self. Additionally, direct sensual
stimulations present during a direct experience can automatically produce affective
responses.
Attitude activation and exemplars. Asch (1940); Lord, Lepper, and Mackie
(1984); and Sia, Lord, Blessum, Ratcliff, and Lepper (1997) conducted experiments in
the areas of attitude activation and exemplars. Asch (1940) asked students to rank the
intelligence, social usefulness, conscientiousness, stability of character and idealism of
ten professions. To influence these rankings, he led some students to believe that most
of their peers had ranked politics last on all five dimensions. Other students were led to
believe that most of their peers had ranked politics first. As predicted, the peer rankings
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
had a large influence on the students’ own rankings. One might conclude that the
students were only conforming to social pressures when ranking the professions.
However, Asch (1940) suggested that since many of the students who ranked politics
first conveyed that they thought about Franklin D. Roosevelt as an example of a
politician and those who ranked politics last thought of Boss Tweed when making their
rankings.
Further, Lord, Lepper, and Mackie (1984) found that students’ attitudes toward
homosexuals predicted their willingness to interact with a homosexual who was more
like the type of person they had in mind when they completed the attitude questionnaire.
These results suggest that attitudes toward social categories might vary when different,
rather than the same, category members come to mind. There is an alternative
explanation for these results and one that supports the social pressure explanation.
When an attitude about a particular group is elicited, the exemplar identified by the
respondent might actually confirm the attitude.
Sia, Lord, Blessum, Ratcliff, and Lepper (1997) gathered information from
students about their attitudes toward politicians, televangelists, talk show hosts, foreign
leaders, homosexuals, rock musicians and stand-up comics. They found that students
who named the same exemplar for a social category on both questionnaires, one month
apart, showed greater attitude stability than students who named different exemplars.
They also found that students who named different exemplars showed greater attitude
instability than those who named the same exemplars. These results suggest that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
exemplar stability is related to attitude stability because different members of a social
category are probably differentially liked.
Sia, Lord, Blessum, Ratcliff, and Lepper (1997) also found that exemplar liking
predicted the direction of spontaneous attitude change from the first administration to
the second. Further, students who were reminded of the same exemplar they had named
previously would retain their earlier attitudes, whereas students who were reminded of a
different exemplar from the initial questionnaire would change their attitudes in the
predicted directions. These results suggest that unstable exemplars can influence
attitudes to become more positive or more negative from one time to the next. If this is
accurate, it follows that social category attitudes can be manipulated by making a more
positive or negative exemplar temporarily accessible. These results may also offer
insight into why attitudes sometimes predict and at other times do not predict behaviors
toward minority groups.
One way to make social category exemplars more salient is through the
frequency of exposure through media coverage. Graham, Weiner, Giuliano, and
Williams (1993) found that public opinion toward people with AIDS improved
immediately after Magic Johnson announced he was HIY positive. It is also important
to research the question, if an exemplar does something to change the person’s opinion
about him or her, does the person change their attitude toward the social category, or do
they retain their initial attitude and subconsciously find an exemplar to support their
initial attitude?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
Werth and Lord (1992) found that people who have an abstraction in mind as
opposed to a specific exemplar for a social category, have attitudes that are more
flexible and susceptible to changing when they have pleasant contact with a social
category member. This combined with Sia, Lord, Blessum, Ratcliff, and Lepper’s
(1997) findings might have implications for attitude change strategies. Identifying
people who do not spontaneously name exemplars may help researchers leam more
about another moderator for the contact hypothesis.
Further, exemplars may influence attitudes similarly to the way generating
reasons do. Wilson, Lisle, and Schooler (1990) found that students who generated
reasons for holding a positive attitude adopted more positive attitudes toward a social
category, and vice versa. It could be hypothesized that people who have positive
exemplars spontaneously come to mind are more likely to adopt more positive attitudes,
whereas people who have more negative exemplars spontaneously come to mind are
likely to adopt more negative attitudes.
Contact with gay men and lesbians. Herek and Capitanio (1996) described three
courses a heterosexual person is likely to follow when they find out someone they know
is gay or lesbian: (a) attach their current anti-gay stereotypes and attitudes to the person
and re-interpret past experiences with the lesbian or gay man to fit the person into their
pre-existing negative attitude schema, (b) maintain positive feelings toward the gay man
or lesbian while viewing them as an atypical example of gay men or lesbians in order to
maintain their existing anti-gay attitudes, or (c) maintain positive feelings toward the
gay man or lesbian and on the basis of those feelings, adjust the categories of lesbians
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
and gay men to become mote persona! and individual. Herek and Capitanio (1996)
assume that the third course of changing long-standing beliefs and strong attitudes
requires the greatest cognitive effort and would be the least likely process chosen.
Although, this assumption requires empirical testing.
Cognitive dissonance. Another possible explanation for the process of how
someone handles the discrepancy between their attitudes and feelings when someone
discloses they are gay or lesbian is through cognitive dissonance. Cognitive dissonance
theory first proposed by Festinger (1957) states that people attempt to maximize their
sense of internal psychological consistency through attitudes, beliefs and values.
Inconsistency in how a person thinks leads to a state of confusion and discomfort,
creating cognitive dissonance. Explicit memory is implicated in this process at the
point when an attribution of the psychological discomfort is made between the initial
attitude and the counter attitudinal behavior. The greater the dissonance, the more the
motivation to change the attitude. Dissonance is thought to be influenced by the
number and strength of thoughts challenging the attitude and the number and strength of
thoughts that support the attitude. In order to return to a state of homeostasis, the
person may be motivated to reduce the dissonance by: (a) changing the relative
proportions of consistent and inconsistent elements by adding more consistent elements,
(b) deleting existing inconsistent elements or (c) altering the importance of the involved
elements. For example, when someone holds a negative attitude toward a group and
they then develop a close, positive relationship with someone from that group, they
experience discomfort with the mismatch between their positive experience and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
negative attitude. Through dissonance reduction processes, the person may adjust his or
her attitude to more closely resemble their experience.
Yeb (1997) provided an illustrative example of this process as it relates to
homosexuality. Jane is homophobic, but she likes Kara who is openly lesbian. Jane
experiences dissonance. If Jane has more consistent elements to support her
homophobic attitude (Jane believes that homosexuality is wrong, Jane is out to
“convert” her) than inconsistent ones (Kara would make a great friend), then she might
choose not to pursue a friendship with Kara. On the other hand, if Jane believes that
Kara is trustworthy, nice and supportive, she might strengthen her thought that the bond
she has with Kara is more important than Kara’s affectional orientation and choose to
weaken her negative attitudes toward homosexuality and continue her friendship with
Kara.
Bern’s (1965) self-perception theory, an alternative theory to cognitive
dissonance, states that people infer their own attitudes the same way they infer the
attitudes of others. They do this by observing their own behavior. As is true in
cognitive dissonance, explicit memory is important in this process. For example,
Brehm’s (1956) free-choice paradigm had female subjects rate how much they liked
each of eight different appliances. The subjects were then asked to choose which of two
appliances she preferred to take home as compensation for participation. After
choosing between the two appliances they had rated almost equally, they were asked to
rate the eight appliances again. Brehm (1956) found that subjects rated the chosen item
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
higher and the rejected item lower than they had in their original ratings. These results
suggest that the subjects created attitudes to justify their behavior.
Contact hypothesis and attitudes toward gay men and lesbians. Allport’s (1954)
contact hypothesis states that if one can positively alter beliefs about a certain group
through education or a positive experience with an individual in the group, then
discrimination will decrease. An extensive body of research about the contact
hypothesis has focused on readily obvious stigmas such as race and physical disability,
while there has been little focus on concealable stigmas such as religion and sexual
orientation.
Herek and Capitanio (1996) examined the relationship between majority group
attitudes and contact with gay men and lesbians. Many gay men and lesbians pass as
heterosexual at least some of the time. Therefore, heterosexuals often learn that a
friend, co-worker or family member is gay or lesbian after they have known the person
for a long time, through either the gay or lesbian person him or her self or through a
third party.
There have been numerous studies conducted with convenience samples (Herek,
1988) suggesting that heterosexuals who know someone who is gay or lesbian,
generally hold more favorable attitudes toward gay men and lesbians. Using a random
sample, The Henry J. Kaiser Family Foundation (2001) found that 47% of those who
know someone who is gay, lesbian or bisexual believe that homosexual behavior is
morally wrong, while a significantly higher percentage (68%) who do not know a
lesbian, gay or bisexual person believes the behavior is morally wrong.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Herek and Capitanio (1996) hypothesized that contact experiences with two or
more gay or lesbian individuals would be associated with more favorable attitudes than
contact experiences with only one person. Second, they hypothesized that contact with
close friends or immediate family members who are gay or lesbian would produce more
positive attitudes than contact with acquaintances or distant relatives who are gay or
lesbian. They also predicted that heterosexuals who have been told directly by a gay
man or lesbian about their sexual orientation would hold more positive attitudes toward
gay men and lesbians based on past research suggesting that self-disclosure of personal
information generally leads to greater liking of an individual. Due to a lack of
knowledge about lesbians and gay men and the stigmatization of homosexuality, they
hypothesized that most respondents would hold more negative than favorable attitudes
toward gay men and lesbians.
Herek and Capitanio (1996) conducted a telephone survey of randomly chosen
English-speaking adults in the 48 contiguous states. They found that female
respondents were more likely than male respondents to know a gay man or lesbian.
Respondents were more likely to describe contact with gay or lesbian friends or
acquaintances (75%) than with relatives (22%). Contact with a gay man or lesbian was
predicted by being female, having a higher educational level, not attending religious
services frequently, being younger, living in one of the Pacific coast states and having a
higher income. Of those respondents who knew a gay man or lesbian, those who were
told directly about the person’s sexual orientation were more likely than others to be
politically liberal, single and urban residents. A majority of the respondents expressed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
negative attitudes toward gay men and lesbians (e.g., “I think lesbians are disgusting.”).
Also as predicted, although knowing one gay man or lesbian was associated with more
positive attitudes than knowing none, knowing two or more gay men and lesbians
generally led to significantly more positive attitudes. Having a gay or lesbian friend
was almost always associated with direct disclosure and more positive attitudes, than
heterosexuals who knew a gay or lesbian person as an acquaintance or distant relative,
and found out about the person’s sexual orientation through a third party.
These results are consistent with the contact hypothesis. Interpersonal
relationships characterized by intimacy, shared values and common goals are more
likely than superficial or distant relationships to be associated with more positive
attitudes toward gay men and lesbians.
Herek and Capitanio (1996) also suggested that since sexual orientation is most
times a concealable stigma, this allows for personal disclosure, which appears to lead to
more significant attitude change than if a person finds out from a third party or never
learns about a friend or co-worker’s homosexuality. They concluded that one of the
most important and effective strategies for reducing prejudice and negative attitudes
toward gay men and lesbians is for each gay man and lesbian to come out to those
closest to them. In terms of public policy, these findings provide a strong argument for
why the U.S. military policy of “don’t ask, don’t tell” is detrimental to all those who
serve in the armed forces. If lesbian and gay military personnel were able to disclose
their sexual orientation to their fellow service members without fear of being
discharged, attitudes toward gay men and lesbians might positively shift. Herek and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Capitanio (1996) suggested future research should include qualitative designs as well as
longitudinal survey studies of heterosexuals’ attitudes to better understand the causal
relationships between contact and attitude change.
Crawford and Solliday (1996) found that their sample of undergraduate college
students consistently held negative attitudes toward gay male and lesbian parenting.
Subjects described gay male and lesbian couples as less emotionally stable than their
heterosexual counterparts, having poor potential to be parents and unable to provide a
loving home for a child. These results suggest that biases and prejudice remain intact
within American culture regarding gay male and lesbian parenting. Similarly, The
Henry J. Kaiser Family Foundation (2001) found that less than half (46%) of the
general public supports adoption rights for gay and lesbian couples. Approximately the
same number (47%) oppose this policy. More than half (56%) of those surveyed
believe that gay and lesbian couples can be just as good parents as heterosexual couples,
while 39% disagree with this statement.
The contact hypothesis suggests that a sound attitude change strategy for gay
and lesbian parents is to get involved in parents’ groups. Heterosexual parents’
negative perceptions of gay and lesbian parents will most likely be challenged if the
lesbian and gay parents hold similar values as the heterosexual parents. Parents who
have negative attitudes toward gay and lesbian parenting and who encounter a gay or
lesbian parent who is similar to them in other ways, beside parenting, are likely to be
subject to cognitive dissonance. If a relationship forms between the gay and lesbian
parents and heterosexual parents who have negative attitudes, their view about gay and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
lesbian parenting is likely to shift toward the positive allowing the heterosexual parents
to have consistency in their attitudes and affectional objects. This proposed theory for
attitude change toward gay and lesbian parenting requires empirical testing.
The extended-contact hypothesis states a person with an in-group friend
(belongs to the majority group) who has an out-group friend (belongs to a minority
group), should have some effect on the person’s attitude toward the out-group. Miller
and Brewer (1986) suggested that the in-group/out-group friendship indirectly
influences the person by incorporating the out-group as part of the self. Wright, Aron,
McLaughlin-Volpe, and Ropp (1995) found not only that in-group/out-group
friendships had a positive influence on intergroup relations and attitudes, but also that
the in-group participants without relationships with the out-group showed a decrease in
intergroup hostility, supporting the extended-contact hypothesis described above.
Mackie (1998) suggests that the important parallels between close relationships, group
membership and attitudes toward groups require further research.
Summary of influence of contact hypothesis on attitudes toward gay men and
lesbians. Evidence suggests that contact experiences with two or more gay or lesbian
individuals is associated with more favorable attitudes than contact experiences with
only one or no gay men or lesbians. Contact with close friends or immediate family
members who are gay or lesbian tend to produce more positive attitudes than contact
with acquaintances or distant relatives who are gay or lesbian. No studies examined if
there were gender differences between being close to gay men and lesbians as separate
groups. Heterosexuals who have been told directly by a gay or lesbian person about
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
their sexual orientation tend to hold more positive attitudes toward gay men and
lesbians. Bern’s (1965) self-perception theory, people create attitudes to justify their
behavior (e.g., intimacy with best friend who discloses he or she is gay or lesbian), is
one explanation for this attitude formation and change process. Another similar
explanation is Festinger’s (1957) cognitive dissonance theory.
Review o f the Literature on Media Effects
“America’s media managers create, process, refine and preside over the
circulation of images and information which determine our beliefs and attitudes and,
ultimately, our behavior” (Schiller, 1973, p.l). This quote begs the question, does mass
media merely reflect reality or does it create and shape it? It’s probably a combination
of reflection and creation. Gross (1991) suggested that the mass media, especially
television, have become a main source of common information and images that help to
create and maintain people’s worldview and value system. Supporting this view, The
Henry J. Kaiser Family Foundation (2001) found that thirty-six percent of those
surveyed stated that media representations have either “a lot” or “some” influence on
their views toward lesbians, gay men and bisexuals (Igb). Only personal experiences
with Igb people ranked higher overall with 49%.
Gross (1991) also stated the mass media plays a major role in projecting images
of minority groups about which there is limited or no opportunity for first-hand
experience and learning (e.g., personal relationships with self-identified gay men and
lesbians), and is rarely positive. Many times these images have been narrow and
negative stereotypes of which most people have little choice other than to accept as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
representative of most minorities. For example, out of 32 films with prominent lesbian
and gay characters from 1961 through 1976, 13 feature gay or lesbian characters who
commit suicide and 18 have the gay or lesbian character murdered (Russo, 1987).
Nardi (1997) suggests that the primary reasons for negative or no portrayals of
lesbians and gay men in entertainment media, especially in wide release films, is
connected to economic, political and social forces. Content has typically been dictated
by what will increase viewer ship. However, with the proliferation of cable television,
there is a new opportunity for expanded images of minorities.
Although there is little specific empirical evidence confirming that minority
issues portrayed by the media, particularly news media, have considerable impact on
public perceptions, there is evidence of media effects on other issues such as violence
and aggression.
An important psychological concept to consider when examining media effects
is called priming effects. Simply stated, priming effects occur when a specific stimulus
is correlated with specific behaviors that follow. For example, the most widely studied
area of priming effects is in the area of television and movie violence. Many studies
have shown that viewing aggression in the media increases the chances that people in
the audience will act aggressively themselves (Andison, 1977; Milavsky, Kessler, Stipp,
& Rubens, 1982). It follows that violence against minorities in the media may incite
violence toward minorities who are negatively portrayed by the media. To create new
and more realistic depictions of minorities, writers and producers can use the priming
effect. For example, images of families and friends displaying unconditional love
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
toward lesbian and gay people may prime people to act positively toward gay and
lesbian people they encounter in their everyday lives. Or, showing more Blacks in
professions like finance, medicine and business can help dispel the stereotype that
Blacks are either professional athletes or drag dealers. This strategy might also
influence and shift norms toward acceptance and fairness.
In addition, Gross (1984) found that heavier television viewing (defined as four
or more hours a day) was related to negative attitudes toward homosexuality. These
results held even after controlling for age, religion, education, race and residence. Even
though those with a higher level of education tended to hold more positive attitudes
toward homosexuality, there was no significant difference between heavy-viewing
higher educated people and heavy-viewing lower educated people.
Many of the inclusive entertainment media depictions of gay men and lesbians
during the 1990s to the present have been non-sexualized lesbian or gay characters that
do not pose a threat to many heterosexuals (e.g., Will & Grace). Most of the more
accurately portrayed images of lesbian and gay men have been of White, middle or
upper class, gay men.
Since the 1970s and 1980s, non-profit organizations such as the Gay & Lesbian
Alliance Against Defamation (GLAAD) in the 1980s to present, NGTF (National Gay
Task Force) in the 1970s and Gay Activist Alliance in the 1970s have been
campaigning for improved representation of lesbian, gay, bisexual and transgender
people in all forms of media. For example, when one sitcom writer at a meeting with a
GLAAD representative stated that he did not know how to communicate quickly to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
audience that his character was gay, except by using stereotypes (more effeminate voice
or mannerisms), it was suggested that he review the ways he communicates how
characters are heterosexual. Additionally, non-reporting of major lesbian and gay
events distorts how lesbians and gay men are viewed. For example, the 1994 Gay
Games in New York, was ignored in almost all sports coverage even though Gay
Games is the largest amateur sporting event in the world. More recently, during the
news coverage of the terrorist attacks on September 11, 2001 the fact that Mark
Bingham was gay, was omitted in almost all media coverage about the five “heroes” of
Flight 93 who assailed the terrorists leading to the jet liner’s crash in Pennsylvania.
Mark Bingham was identified as one of the probable heroes and it took GLAAD and
others to convince some journalists that his sexual orientation was an important part of
his identity and that Americans need to see images like Mark Bingham’s to challenge
their beliefs about gay men and lesbians.
Williams (1997) suggests “a political approach cannot be effective without
addressing prejudicial attitudes and institutionalized discrimination in the general
population” (p. 3). He concludes that with such low public acceptance of gay men and
lesbians in America, maybe the focus should be on changing attitudes. Due to the
dearth of research on the effects of the media on attitudes toward gay men, lesbians,
women and racial minorities, it is evident there needs to be research done specifically
on the effects of all forms of media (print, television, radio, wire services, Internet,
video games and film) on attitudes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
Summary o f media influences on attitudes toward gay men and lesbians. One
study showed a relationship between heavy television viewing (greater than four hours
per day) and more unfavorable attitudes toward gay men and lesbians. A more recent
study using a probability sample found that fully 36% of heterosexuals surveyed stated
that media representations of gay men and lesbians had “a lot” or “some” influence on
their attitudes toward gay men and lesbians. Only personal experiences with gay men
or lesbians ranked higher than media images when it came to identifying influences on
respondents’ attitudes toward gay men and lesbians.
Review o f the Literature on Attitude Function Research
In an extensive literature review on attitude change, Wood (2000) concluded
that attitude change can be motivated by a variety of normative concerns that are linked
to the function or functions of a particular attitude. Agreeing with others in order to be
accepted by them or to improve relations with others is one example of a normative
concern that is directly linked to the social-expressive attitude function (e.g., the reward
is acceptance, the punishment is rejection). The normative concern expressed by the
defensive and value-expressive functions is linked to the motivation to ensure the
integration and favorable evaluation of the self. The normative concern of holding
accurate or valid attitudes is expressed through the experiential-schematic or knowledge
function.
Haddock, Zanna, and Esses (1993) studied the impact of stereotypes, affect
(emotions in response to interactions with homosexuals), past experiences with gay men
and lesbians and perceived value dissimilarities on attitudes toward gay men and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
lesbians. Using the Right-Wing Authoritarian (RWA) scale, they also analyzed the
mediating effects of authoritarianism on these variables. Overall, more negative
attitudes were correlated with higher authoritarianism scores. In addition, the predictive
variables differed for high and low authoritarians. From their results, Haddock, Zanna,
and Esses (1993) concluded that high authoritarians tend to place great importance on
the values held by gay men and lesbians, on differences between their own values and
perceived values of gay men and lesbians and on their past experiences with gay men
and lesbians when forming their attitudes toward gay men and lesbians.
All of these factors seem to fit into the four functions of attitudes—experiential-
schematic; value-expressive (perceived difference in values, customs and traditions
denoting good and bad or right and wrong); defensive (perceived threat by the out
group); and social-expressive (negative attitudes toward homosexuality is the norm for
authoritarians). From their results, Haddock, Zanna, and Esses (1993) concluded that
educational programs would be most effective in positively influencing low
authoritarians. Whereas, changing the negative attitudes of high authoritarians, would
require strategies aimed at reducing perceived differences in values.
Luhrs, Crawford, and Goldberg (1991) conducted a study to isolate and examine
the presence of the defensive attitudinal function as a predictor of affective responses of
college students toward gay men and lesbians. The intervention message was a
videotape of excerpts from the film, The Times of Harvey Milk. This documentary was
chosen to be the independent variable because the investigators determined that it
provides accurate information about homosexuality and lesbians and gay men. They
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
claimed the film also dispels popular stereotypes of lesbians and gay men and they felt
it would arouse emotion and awareness of how negative attitudes can translate into
painful outcomes for lesbians and gay men and those who care about them. They found
that when the defensive function was present in individuals, they exhibited a
significantly higher level of negative attitudes than those subjects who scored low on
the defensive measure. Other significant findings included defensive function present
individuals were more likely to be male, with lower levels of education, young and
single. Additionally, males were found to hold more negative attitudes overall than
females and were also more likely to hold more negative attitudes toward gay men than
toward lesbians.
After viewing the videotape, all subjects reported significantly less negative
attitudes toward gay men and lesbians. These results suggest that individuals who hold
negative attitudes toward gay men and lesbians primarily expressed as a defensive
function, may not be as resistant to attitude change as previously theorized. More than
50% of the individuals surveyed did not have the defensive function present. These
individuals reported lower homophobia scores than those with the defensive function
and they were more likely to report past positive direct experiences with gay men and
lesbians. Since this experimental design did not include a follow-up study to determine
if the attitude change was long lasting, more research is needed to test this interpretation
of the results. Additionally, once participants realized the interview was aimed at
changing attitudes, they might have responded in a socially desirable manner.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
Kristiansen (1990) found that feminist lesbians (primarily involved in the
women’s rights movement) had less favorable attitudes toward gay men, associated
with fewer gay men, perceived less common fate with gay men, wished to cooperate
less with gay men and perceived less value similarity with gay men than did gay
movement lesbians (primarily involved in the gay rights movement). From their
findings, they concluded that feminist lesbians shared an intergroup relationship with
gay men, whereas gay movement lesbians and gay men shared an intragroup
relationship. Consistent with the value-expressive function, feminist lesbians’ attitudes
toward gay men were explained by their perceptions that gay men placed less
importance on values they themselves regarded as important. These findings support
the social identity theory’s claim that groups will underestimate the similarity between
their own group and members of an out-group to achieve a distinct and positive social
identity. Feminist lesbians perceived absolutely no similarity between their own values
and those of gay men, when in fact, the gay men and feminist lesbians surveyed shared
many of the same values. Further, all three groups underestimated the actual degree of
value similarity between the other groups, but feminist lesbians did so at a much higher
rate than did the other two groups. Additionally, these findings support the claim that
while many people appeal to values to justify their attitudes toward social issues and
groups, they may also exaggerate perceptions of value differences between their group
and the out-group to strengthen their social identity.
Kristiansen and Zanna (1988) examined Eiser’s (1987) value-justification
hypothesis, which states that individuals with opposing attitudes towards an issue hold
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
different values to explain their attitudes. Eiser (1987) posited that not only would
individuals appeal to different values to explain their attitudes, but also they would
argue that the values they used to justify their position are more relevant to the issue at
hand. For example, someone may believe that individual freedom is more relevant
when explaining his or her pro-choice position on abortion. Whereas, a pro-life
individual might state that the value of human life is the most relevant value to consider.
Kristiansen and Zanna (1988) hypothesized that subjects with opposing attitudes toward
each attitude object would differ in the values they identified as relevant in evaluating
the issue.
Kristiansen and Zanna (1988) hypothesized that low self-monitors (guided by
their inner attitudes and predispositions and don’t tend to make evaluations based on
what others might think about them) would utilize widely held values to justify their
attitudes, thus fulfilling a value-expressive attitude function. In contrast, high self
monitors (attitudes are guided by situational and external cues such as peer pressure)
would be expected to fulfill a social-expressive attitude function.
Using a convenience sample, Kristiansen and Zanna (1988) instructed subjects
to rank the order of importance and relevance of 18 values (Rokeach, 1967) as they
related to their attitudes toward abortion and nuclear weapons. As hypothesized,
subjects with opposing attitudes identified different values as relevant to the issue. For
example, subjects who were in favor of allowing nuclear weapons in Canada regarded
‘national security’ as more relevant in comparison to subjects opposed to nuclear
weapons who viewed ‘wisdom’ as more relevant value to consider. Even when value
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
importance was controlled for, value relevance effects were still significant. The results
also supported their second hypothesis. As expected, low self-monitors’ perceptions of
value relevance were significantly different for those with negative versus positive
attitudes toward abortion. In contrast, high self-monitors did not show this effect,
providing support for the hypothesis that low self-monitors would show value-
justification effects, while high self-monitors would not. Since the value-justification
hypothesis held across two relatively independent attitude objects, Kristiansen and
Zanna (1988) suggested this effect might generalize to other important social issues.
Maio and Olson (1994) attempted to determine whether attitude functions
moderate the relation between value importance and attitudes. They did this by
comparing the value-expressive function and the utilitarian function (attitudes are held
to maximize rewards and minimize punishments obtained from the environment). They
chose the utilitarian function because they assessed that it was the least related function
to the value function. They used Shavitt’s (1989) thought-listing technique to assess
attitude functions.
Maio and Olson (1994) were also interested in determining whether attitude
functions moderate the relation between value importance and behavior. They assessed
value importance and functions of a benefit dance to support the construction of an
enclosed smoking area on a college campus. As expected, they found that attitude
functions did moderate the relation between value importance and attitudes. Subjects
with value-expressive attitudes showed a stronger relationship between values and
attitudes as well as between value importance and attitudes as opposed to those subjects
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
with either utilitarian attitude or mixed attitude functions. Overall, there was no
significant relationship between values and attitudes, supporting the idea that many
people do not form attitudes that are expressive of their values.
Matching versus non-matching functional approach to attitude change. To
further the understanding of the links between values, attitudes and behaviors, Murray,
Haddock, and Zanna (1996) created an experimental approach to produce value-
expressive attitudes. Then, they examined whether value-expressive attitudes change
most in response to arguments focusing on values and if adversity increases an
individual’s commitment to value-expressive attitudes and lowers an individual’s
commitment to social-expressive attitudes. Murray, Haddock, and Zanna (1996)
identified a need for experimental research of this nature as they identified that most of
the research examining the relation between holding particular values and attitudes has
been correlational in nature.
Murray, Haddock, and Zanna (1996) incorporated Ostrom and Brock’s (1969)
values-bonding technique in order to link a particular attitude to a subject’s central
values. This procedure has been shown to create a condition whereby subjects’
attitudes become more resistant to change. In a series of three experiments, Murray,
Haddock, and Zanna (1996) first convinced subjects that they were the type of person
whose attitudes typically fulfilled either a value-expressive (YE) or social-expressive
(SE) function. They were successful in creating this manipulation. Replicating the
findings of DeBono (1987) and using institutionalization of mentally ill people as the
attitude object, they found that high self-monitors (those fulfilling an SE function)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
changed their attitudes most significantly in response to arguments that linked the
particular positions to the beliefs of important reference groups. In contrast, they found
that low self-monitors (those more likely to fulfill a VE function based on internal
monitoring) changed their attitudes more significantly in response to arguments that
linked particular positions to subjects’ personal values. These results lend support to
the functionality-persuasion hypothesis stating that attitudes change most in response to
arguments that address and match the function of the attitude.
In order to test the conclusion presented in Johnson and Eagly’s (1989) meta-
analytical literature review of the persuasion literature, Murray, Haddock, and Zanna
(1996) chose to focus on two findings: (a) individuals changed their attitudes less in
response to persuasive arguments when their attitudes reflected their values and (b)
individuals who experienced adversity as a result of their attitudes, showed an increased
commitment to those attitudes. Using affirmative action as the attitude object, Murray,
Haddock, and Zanna (1996) predicted that individuals holding VE attitudes and suffered
as a result of those attitudes would show an increased commitment to those attitudes.
They suggested that VE individuals who held on to value-relevant attitudes in the face
of adversity would be acting with integrity and preserving their sense of self-esteem.
They also suggested that SE individuals might not be as strongly motivated by self
esteem when faced with adversity because of an attitude they hold. Therefore, they
predicted that SE individuals might be more likely to abandon SE attitudes when they
suffer because of them. As predicted, Murray, Haddock, and Zanna (1996) found that
individuals in the VE condition became more committed to affirmative action when
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
they suffered because of their ideology. In contrast, they found that individuals in the
SE condition became less committed to affirmative action when they suffered because
of their ideology.
Edwards (in press) also found that using in order to maximize attitude change,
the persuasive message should match the attitude function (psychological motivation).
These results are important because they provide information about possible attitude
change strategies to be used with people who hold value-expressive versus social-
expressive attitudes. Those who hold strong value-expressive attitudes may need to
hear arguments based on their values in order to change their attitudes. Whereas, those
who hold social-expressive attitudes may need to receive pressure from their reference
group to change their attitudes to conform to the group’s dominant attitude. These
results also imply that value-expressive attitudes may be strengthened as a result of
adversity (e.g., American male who holds unfavorable attitudes toward gay men and
lesbians motivated by his religious beliefs may feel stronger about this if his house of
worship experiences adversity such as picketing by gay rights activists). On the other
hand, someone who holds unfavorable attitudes toward gay men and lesbians motivated
by social group norms may be more apt to adopt a more favorable attitude toward gay
men and lesbians when holding these attitudes creates psychological conflict (e.g., a
gang member who learned that the group norm is a hatred and disgust toward gay men,
who then goes to prison as a result of a gay-bashing he was involved in).
In contrast to the matching approach tested by Murray, Haddock, and Zanna
(1996), Millar and Millar (1990) examined attitude change as a function of attitude type
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
and argument type and found differing results. They hypothesized that affective-based
attitudes would be more likely to change in response to rational arguments and
cognitive-based attitudes would be more susceptible to change in response to emotional
arguments. They pointed out that there have been few attempts to combine the attitude
type and argument type to determine what combinations affect attitude change. Millar
and Millar (1990) proposed that when the argument and the attitude are based on the
same class of information (rational-cognitive or emotional-affective), the argument
threatens the way in which the individual has thought about the object and thus the
person will be motivated to create counter-arguments to ultimately reject the argument.
On the other hand, when the counter attitudinal argument and attitude are based on
different classes of information (rational-affective or emotional-cognitive), the
argument will not directly threaten the way in which the individual has thought about
the object. For example, if a person dislikes gay men and lesbians because of intense
negative emotions associated with them, a rational argument promoting gay and lesbian
rights (e.g., lower government spending on dealing with the high prevalence of drug and
alcohol addiction linked to stigmatization and criminalization of homosexuality in many
states), does not suggest that homosexuality promotes positive feelings or that the
person has failed to consider other emotions evoked by homosexuality. In this
situation, there is relatively little threat to the way in which the person has chosen to
think about the object and little motivation for counter arguing.
As hypothesized, Millar and Millar (1990) found that affectively based and
cognitively based attitudes were changed by arguments emphasizing a component
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
(affective or cognitive) different from the component on which the attitude was based.
Millar and Millar (1990) believed that the different arguments used in their studies
manipulated emotion and cognition and did not manipulate novelty. If novelty was
manipulated, it has been shown that presenting unique or novel information creates
more persuasive arguments than when common or widely known information is
presented. Millar and Millar (1990) stated they were confident that the experimental
design controlled for novelty.
Summary of attitude Junction research. There is evidence to suggest that less
favorable attitudes toward gay men and lesbians are associated with individuals who
have the defensive and the value-expressive functions dominant compared with
individuals with a dominant social-expressive or experiential-schematic function. There
is some evidence to suggest that men who hold a dominant defensive function hold
more negative attitudes toward gay men and lesbians than women who hold a dominant
defensive function. One study showed that 10% of their subjects predominantly held
the defensive function for attitudes toward gay people. Another study showed over
60% of the sample manifested a social-expressive function for attitudes toward gay men
and lesbians.
All of the studies examining these relationships used college-aged convenience
samples, which might give a good explanation for the high occurrence of those
motivated to hold their attitudes toward gay people for social reasons—peer pressure in
college motivating their beliefs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
Additionally, there appears to be conflicting evidence when testing a functional
matching and mismatching approach to attitude change. More research needs to be
conducted to confirm or disconfirm the validity of these attitude change theories.
Review o f the Literature on Gender Differences and Item Order
There are consistent findings that heterosexual women generally hold more
favorable attitudes toward homosexuals than their male counterparts with a few
exceptions. On the other hand, the ability to analyze the gender gap by attitude target
has not been possible due to polling and survey designs. Over the past 40 years, most
studies examining attitudes about sexual orientation have measured attitudes toward
homosexuals as one group, assuming gay men and lesbians are one distinct attitude
object. Recently, Herek (2002) demonstrated the importance of differentiating lesbians
and gay men as attitude targets in survey research—there are most likely different
schemas activated when a heterosexual man or woman is asked about a gay man or
about a lesbian.
Stevenson and Gajarsky (1990) found that heterosexual women and men did not
differ significantly in their attitudes toward lesbians and gay men, but attitudes toward
lesbians were more easily changed than attitudes toward gay men regardless of the
gender of the respondent. The finding that attitudes toward lesbians were more pliable
than attitudes toward gay men could be important to consider when creating strategies
for visibility and attitude change.
Stevenson and Gajarsky’s (1990) finding that women and men did not
significantly differ in their attitudes toward lesbians and gay men contradicts the finding
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
by Kite (1984) that suggested men ate somewhat less tolerant of homosexuality than
women. Oliver and Hyde (1993) conducted a meta-analysis on gender differences in
sexuality from the 1960s through the 1980s. Like Stevenson and Gajarsky (1990), they
found there were no gender differences in attitudes toward homosexuality and no
gender differences in attitudes about civil liberties for gay men and lesbians. In
response to Oliver and Hyde (1993), Whitley and Kite (1995) critiqued the meta
analysis and ultimately conducted their own meta-analysis. Whitley and Kite (1995)
used 66 studies providing 107 effect sizes, while Oliver and Hyde’s (1993) sample only
included 16 studies with 42 effect sizes. Whitley and Kite (1995) found that men
generally held more negative attitudes toward homosexuality and toward the civil rights
of lesbians and gay men than did women. They also found that effect sizes were
moderated by age. Adult samples showed no sex differences and college aged samples
showed larger effect sizes.
Social role theory and numerous other studies (e.g., Herek & Capitanio, 1996;
The Henry J. Kaiser Family Foundation, 2001; Herek, 2002) tend to support the
findings of Whitley and Kite (1995) and Kite (1984). Social role theory assumes that
heterosexuality is part of both the male role and the female role (Bern, 1981). Deaux
and Lewis (1984) found that males who have a large degree of feminine qualities were
assessed by heterosexuals as having a higher probability of being gay than are men
described as having masculine qualities. On the other hand, females who have a high
degree of masculine qualities were assessed by heterosexuals as having a much lower
probability of being a lesbian. These findings suggest that role violations for males are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
more serious than for females. It predicts that homosexuality will be viewed as a more
serious violation of roles by males than by females resulting in males holding more
negative attitudes.
Herek (2000) took a unique look at gender differences in attitudes toward gay
men and lesbians. Herek (2000) identified two approaches used in the gay and lesbian
movement to change the status of gay and lesbian people in society. Each of these
approaches has its own paradigm for understanding the roots of sexual prejudice. The
dominant approach today is based on the minority group politics model and focuses on
securing civil and human rights for gay and lesbian people that are equivalent to those
of heterosexuals. This strategy emphasizes the reality that gay men and lesbians have
visible communities with cultural traditions, political interests and other commonalities.
Therefore, gay men and lesbians constitute what is very similar to an ethnic group. This
gay rights framework suggests that attitudes toward gay and lesbian people are
psychologically similar to majority attitudes toward racial, ethnic and other minority
groups.
The second strategy, gay liberation, popular in the 1960s and early 1970s, had
the goal of altering society’s view of sexuality and gender. This strategy sought to
unleash the bisexual potential in everyone, eliminating the need for the categories of
homosexual and heterosexual.
This strategy differs from gay rights because it views sexual prejudice as
specifically linked to attitudes toward two aspects of personal identity. Whereas, gay
rights are specifically linked to attitudes toward a quasi-ethnic group. The first aspect
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
of personal identity addressed by the gay liberation paradigm is about one’s attitudes
toward their own sexuality in terms of where they view themselves on the homosexual-
heterosexual continuum. The second personal identity aspect addressed by gay
liberation is how sexual prejudice is linked to attitudes toward one’s own gender
identity. Expressions of sexual prejudice can help a person to demonstrate that he or
she is heterosexual, but also that he or she understands and adheres to prescribed gender
roles. For men in particular, there is a lot of pressure to follow male gender role norms
and to reject gay men in order to affirm their heterosexual masculinity.
Herek (2000) hypothesized that attitudes toward gay men and lesbians should
not differ when evaluations are made based on minority group membership, regardless
if there are normative pressures from one’s peers (social-expressive function), political
or religious values (value-expressive function) or intergroup contact (e.g., experiential-
schematic function). Herek also hypothesized that attitudes toward gay men and
lesbians should differ between men and women when evaluations are made based on
the attitudes toward one’s sexual or gender identity. He suggested that when the
function of an attitude is to demonstrate one’s masculinity or femininity, and in turn,
their heterosexuality, sexual prejudice will most likely be directed at the subgroup of
homosexuals most closely tied to their own sexual and gender identity, homosexuals of
one’s own sex. The stronger negative evaluation of homosexuals of one’s own gender
is motivated by the desire not to be confused as belonging to this particular group.
There is a much greater concern for men than for women in American society to
demonstrate one’s heterosexuality, and, at the same time, one’s gender role conformity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
Therefore, Herek (2000) hypothesized that differences in heterosexuals’ attitudes
toward lesbians and gay men should result primarily from differences in how
heterosexual men view lesbians compared to gay men. Herek (2000) predicted there
would only be a stronger negative evaluation of homosexuals of one’s own gender for
heterosexual men, as compared to heterosexual women.
Most of the data on attitudes toward gay men and lesbians has been based on
surveys done with samples of college students. The limitations from these
nonprobability student samples are heightened by the fact that sexual prejudice is
correlated with educational level and age (Herek, 1984). In their meta-analysis, Kite
and Whitley (1996) found that national surveys with probability samples have mostly
focused on opinions about civil liberties for gay men and lesbians, whereas convenience
samples of college students have focused on affective responses to homosexual
behaviors or to gay men and lesbians as people, without asking respondents about gay
men and lesbians separately. Sex differences were found depending on the type of
attitude being assessed as well as the type of sample used. The most significant sex
differences were found with college students and in personal responses to gay men and
lesbians as people. Additionally, there were pronounced sex differences in probability
samples about gay parenting and military service.
Herek (2000) concluded that what was needed were responses from a
probability sample of heterosexuals to items eliciting affective and cognitive responses
to gay men and lesbians and to distinguish between the two groups. Additionally,
Herek determined it was important to determine if the order in which items were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
asked—questions about gay men versus those about lesbians—would influence
responses. This information is important because it could shed light on the effects that
one attitude object (e.g., lesbians) may have on another related attitude object (e.g., gay
men).
Using the Attitudes Toward Lesbians (ATL), Attitudes Toward Gay Men (ATG)
Scales (Herek & Capitanio, 1996, 1999) and feeling thermometers designed to measure
respondents general feelings of emotional warmth or coldness toward gay men and
lesbians, as well as toward certain ethnic and racial groups, a telephone survey of 1,309
English-speaking adults in the United States who were randomly chosen was conducted
by Herek in 1997.
Significant results from this study included: (a) heterosexual men’s ATG scores
were significantly higher than their ATL scores, (b) to a lesser extent heterosexual
women’s ATL scores were higher than their ATG scores, (c) both men and women
expressed more coldness toward gay men than toward lesbians, (d) heterosexual Blacks
showed similar context effects and held similar attitudes as their White counterparts and
(e) mean ratings of warmth toward Whites, Blacks and Mexican-Americans were higher
than those toward gay men and lesbians.
Herek (2000) concluded that there are clear sex differences in attitudes toward
gay men and lesbians. Consistent with most data from convenience samples, responses
from this random sample indicated that heterosexual men’s attitudes differed
significantly according to the gender of the attitude target. Namely, their attitudes
toward gay men were significantly less favorable than their attitudes toward lesbians.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
He also concluded that heterosexual women’s attitudes did not differ reliably according
to target gender of the attitude object. Further, examination of specific responses to the
ATG and ATL scale items revealed that most men and women found male and female
homosexuality to be wrong and unnatural; and most men and not a majority of women,
expressed disgust at both lesbians and gay men. Overall, attitudes toward both lesbians
and gay men were generally negative and highly correlated.
Item order differences. Significant differences were found when analyzing the
order of survey items. Men expressed less prejudice toward lesbians when the lesbian
items were presented before the gay male items. Interestingly, men’s ATG scores,
women’s ATG scores and women’s ATL scores were not significantly different when
analyzing item order. However, when heterosexual men were asked questions first
about gay men and then about lesbians, their attitudes toward both groups were
considerably more negative than when the lesbian items came first. Therefore,
contextual cues may play a significant role in how heterosexual men express their
attitudes toward lesbians. This is important because a significant amount of attitude
theory literature defines attitudes in terms of long-term memory structures that are
activated when an attitude object is brought to mind. Once a specific attitude is
activated, other related attitudes within the respondent’s cognitive network are also
activated resulting in a spreading effect. Tourangeau and Rasinski (1988) suggested
that being asked questions about one attitude object probably activates certain feelings,
beliefs and evaluations that will influence answers to subsequent questions about a
related attitude object.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
Herek (2000) suggested that when attitudes about gay men are expressed, they
activate different associative networks from those activated by questions about lesbians.
This held true for heterosexual men who when expressing attitudes toward gay men,
negative thoughts and feelings were retrieved from memory, which influenced their
responses to items about lesbians. Consequently, when men were asked questions about
gay men first and then about lesbians, their attitudes toward both groups were
significantly more negative than when they were asked about their attitudes toward
lesbians first. In other words, when heterosexual men were asked about their attitudes
about lesbians first, their attitudes toward lesbians were significantly more favorable
than when they were asked about their attitudes toward gay men first. There were only
minimal context effects for heterosexual women.
Herek (2000) suggested that an explanation for these findings might be
connected to the distinction between minority group attitudes and gender and sexual
identity. Questions about gay men and lesbians most likely evoke different responses
from heterosexual men and women based on the activation of heterosexual masculine
and feminine roles and sexual identity as well as views about minority groups in
general. Due to pressure in American society for males to identify and act masculine
and heterosexual, rejection of gay men and expression of hostility and negative attitudes
toward gay men logically follows. Herek (2000) suggested that when attitudes toward
gay men are activated for heterosexual men, later questions about lesbians are
subsumed in that schema, thus accounting for less favorable attitudes toward lesbians.
Inquiring about attitudes toward lesbians first, however, may activate responses based
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
more in values (religious), non-threatening sexuality (lesbians as desired sexual objects)
or positive gender identity (lesbians are stereotypically viewed as more masculine than
heterosexual women).
Since rejection of lesbians is generally not as strong a part of heterosexual
women’s self-image, their attitudes toward both gay men and lesbians can be explained
better by minority group attitudes. In other words, minority group values are used to
evaluate gay men and lesbians and these do not differentiate the two groups
significantly.
Herek (2000) concluded that most heterosexual men and women utilize both
minority group attitudes as well as sexuality and gender identity schemas when
evaluating their attitudes toward gay men and lesbians. Future research may test
Herek’s theory that heterosexual men tend to give somewhat greater emphasis to issues
of gender and sexual identity, especially when evaluating gay men. Whereas,
heterosexual women tend to primarily consider their minority group values when
evaluating sexual orientation. Additionally, future research may test Herek’s suggestion
that the dominant presence of the defensive function in heterosexual men is related to
item order—the defensive function is correlated with less favorable attitudes toward gay
men and toward lesbians when questions about gay men are asked first.
Herek and Capitanio (1999) suggested an alternative explanation for these
results. In a phone interview situation, heterosexual women may be better able to
access and retrieve their attitudes about gay men and lesbians. This may be true
because heterosexual women may have had more contact experiences with gay men and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
lesbians and may have greater familiarity with lesbians having grown up female. On
the other hand, heterosexual men typically have fewer opportunities for direct contact
with openly gay men and lesbians (Herek & Capitanio, 1996) and may not have thought
extensively about their attitudes toward lesbians, except to the extent that they may find
them sexually interesting. Thus, asking about attitudes toward lesbians first may elicit
neutral or positive evaluations by heterosexual men. This explanation is consistent with
the cognitive activation hypothesis, which suggests that somewhat different images are
activated when respondents are asked about gay men and about lesbians separately.
Depending upon what stimulus is activated first, the initial attitudes prime the
respondent to react in a similar direction, positive or negative, when the second related
attitude object is presented.
Herek and Capitanio (1996) suggested that the cognitive activation hypothesis
does not explain why specific considerations become linked to attitudes in the first
place and why men and women might differ on these linkages. They suggested that the
functional approach to attitudes provides a plausible explanation for these differences.
Using this approach, their results suggest that the psychological needs of heterosexual
men’s attitudes toward gay men and lesbians differ from those of heterosexual women’s
psychological needs. They hypothesized that for heterosexual women, questions about
lesbians and gay men may activate primarily value-expressive, social-expressive or
experiential-schematic functions, none of which would necessarily create a difference
between their attitudes toward gay men and lesbians. For example, a value-expressive
attitude might be based on how the person views the morality of same-sex relations. In
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
contrast, for heterosexual men, the topic of homosexuality often activates considerations
of masculinity, sexuality and personal threat, all factors that contribute to the defensive
function (Herek, 1988).
Using a college sample, Herek (1987) found that defensive attitudes about
homosexuality were generally associated with attitudes toward one’s own sex and were
usually negative. Herek and Capitanio (1999) suggested that assessing attitudes about
gay men might make the defensive function more accessible for male respondents.
They hypothesized that once a defensive function was activated, all aspects of
homosexuality would be negatively evaluated. Assessing attitudes about lesbians first,
may make other attitude functions more salient (e.g., value-expressive or experiential-
schematic), which might be associated with less negative attitudes. However, Herek
and Capitanio (1999) did not assess attitude functions in their study, rendering these
hypotheses in need of empirical verification.
Considering these findings, Herek and Capitanio (1999) suggested that
interventions to reduce sexual prejudice, especially for heterosexual men, might be
more successful if they address attitudes toward lesbians first.
Summary o f gender differences and item order. Although there are conflicting
results, it appears that heterosexual men generally hold less favorable attitudes toward
homosexuality and toward the civil rights of gay men and lesbians than do heterosexual
women. When as separate reference groups, it appears that heterosexual men hold less
favorable attitudes toward gay men than toward lesbians. And, to a lesser extent,
heterosexual women hold less favorable attitudes toward lesbians than toward gay men.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
It also appears that men’s attitudes, especially toward lesbians, appear to shift based on
item order. Questions assessing attitudes about gay men may activate different
associative networks from those activated by questions about lesbians. Heterosexual
men who answered questions about gay men appear to stimulate negative thoughts and
feelings from memory. These unfavorable images of gay men negatively influence
thoughts and feelings about lesbians as assessed by attitude items referring to lesbians.
In other words, when heterosexual men were asked questions about gay men first, their
attitudes about both gay men and lesbians were more unfavorable than when they
answered questions about lesbians first.
Synthesis o f Review of the Literature
An extensive body of research has been done about attitudes toward readily
apparent stigmas such as race and physical ability. A somewhat lesser amount of
research has been conducted on attitudes toward concealable stigmas such as gay men
and lesbians. A few studies have examined context effects on attitudes toward gay men
and toward lesbians by asking questions about each target group—gay men and
lesbians—in a different order. The results of these studies suggest that under
circumstances like priming and social desirability, attitudes may not be stable and
enduring schemas in the brain, but they are sometimes created spontaneously to satisfy
a pressing psychological need.
Only one study, conducted almost 20 years ago, examined attitude functions
toward gay men and lesbians. No studies have examined the functions of attitudes
toward gay men and toward lesbians separately. Further, no studies have examined
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
close relationships with gay men and lesbians as separate targets of focus. Additionally,
no studies have examined the association between context effects, attitudes and attitude
functions.
Individual and demographic differences such as being female, younger,
educated, rarely or never attending religious services, living in urban areas on the
Pacific or Northeast coasts, higher income and significant contact with gay men and
lesbians have typically been associated with more favorable attitudes toward gay men
and lesbians. One study (Gross, 1984) found a positive correlation between the high
frequency and duration of television viewing and unfavorable attitudes toward
homosexuality, while controlling for demographic differences. Another study found a
relationship between unfavorable attitudes and those who adhere to traditional .
ideologies of family and gender and subscribe to an orthodox religious ideology.
By contrast, most of these studies do not assess attitudes toward gay men and
lesbians as separate groups. Gaining a better understanding of differences in attitudes
toward gay men and attitudes toward lesbians seems especially important, as there has
been an implication in most attitude research about sexual orientation and race that the
male and female attitude objects within those groups represent and activate the same
associative cognitive structures. For example, if asked about attitudes toward African
Americans, there is an underlying assumption that the respondent views African
American men and African American women as one group. Herek (2002), Herek and
Capitanio (1999), and Herek and Capitanio (1996) have shed some light on these issues,
both theoretically as well as empirically. These aforementioned studies used an adult
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
probability sample and data was collected in a national telephone survey. Other survey
methods (e.g., paper and pencil or online) should be used to replicate and extend these
findings.
Despite much research on the association between individual differences and
attitudes toward gay men and lesbians, and only a few studies examining the functions
these attitudes serve, there do not appear to be any studies investigating the
relationships between demographic and social differences and attitude functions.
Additionally, there are no studies examining the association between the attitude
functions for lesbians and the attitude functions for gay men and whether these
functions remain stable when attitudes toward gay men are assessed before attitudes
toward lesbians and vice versa.
Herek (2002) suggested that heterosexual men tend to organize their attitudes
toward gay men and lesbians in terms of gender and sexual identity schemas. This
results in different ways of thinking about lesbians and gay men. By contrast,
heterosexual women tend to organize their attitudes toward gay men and lesbians in
terms of a minority group paradigm that does not differentiate between gay men and
lesbians. Herek went on to surmise that attitudes toward gay men and lesbians should
not differ when evaluations are based on minority group membership. Therefore,
attitudes toward gay men and lesbians should not differ when the social-expressive,
experiential-schematic and value-expressive functions are expressed.
On the other hand, attitudes toward gay men and lesbians should differ between
heterosexual men and women when evaluations are made based on one’s own sexual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
orientation or gender identity. This suggests that those who hold a prominent defensive
attitude function will generally hold more negative attitudes toward those of their own
gender.
One possible reason for holding a defensive attitude function is that some
individuals may have experienced same-sex attraction or engaged in homosexual
behavior, which he or she views as unacceptable and a threat to their self-concept.
Because same-sex attraction or sexual behavior is somewhat more acceptable for
women, it is hypothesized that heterosexual women may endorse mixed attitude
functions—the defensive function along with an additional function to describe their
motivation for holding their attitude toward lesbians.
In some cases, the other function may add a favorable component to their
reported attitude, lessening the negative effect the defensive function has on
heterosexual women’s attitudes toward lesbians. There are no studies examining this
theory in the context of attitude functions.
Due to the prevailing negative attitudes toward gay men and lesbians in society,
it can be hypothesized that social-expressive heterosexual men and women will
generally hold more negative attitudes toward gay men and lesbians. Past findings
(e.g., Haddock, Zanna, & Esses, 1993) suggest that in-group members generally
perceive greater value dissimilarity of out-group members, and those who perceive
greater value dissimilarity generally hold more negative attitudes toward an out-group.
Additionally, because overall in-group attitudes toward gay men and lesbians are
negative and the dominant personality characteristic that defines the social-expressive
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
function is a heightened awareness of others5 reactions to oneself and a strong need for
those reactions to be favorable, social-expressive respondents will fit in by endorsing
more negative attitudes.
Although a few studies have examined the role of interpersonal contact with gay
men and lesbians on attitudes toward gay men and lesbians, no studies have confirmed
the assumption that the prevailing motivation behind these more favorable attitudes is
the experiential-schematic function. Herek and Capitanio (1996) found that knowing a
gay man or lesbian was associated with more positive attitudes than knowing none.
However, only respondents who knew at least two gay men or lesbians were
consistently significantly different from those with no contacts. Lack of significance
for respondents who only knew one gay man or lesbian may have been attributable to
the fact that the one experience might have been negative. Or, the motivation behind
their attitude may have not been experiential-schematic, but predominantly one of the
other three functions.
Purpose of the Study
The primary purpose of this study was to examine whether heterosexuals tend to
hold different attitudes toward lesbians as a separate group and toward gay men as a
separate group. Examining the relationship of demographic (e.g., gender) and social
group differences (e.g., # of close relationships to gay men and # of close relationships
to lesbians as separate independent variables) to attitudes toward lesbians and gay men
as well as to attitude functions was also a primary goal of this study. Another important
goal was to determine if attitude functions predict attitudes. A secondary goal was to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
examine if the order that questions about lesbians and questions about gay men were
asked influence attitudes and attitude functions. Assessing overall agreement with
specific public policy issues for gay men and lesbians and whether or not there would
be gender group differences in opinions was also a secondary goal of this study.
Understanding these relationships could prove to be instrumental in creating effective
strategies to combat sexual prejudice.
Research Questions and Hypotheses
Research Question 1: Are there demographic and social group differences between
heterosexuals’ attitudes toward lesbians, attitudes toward gay men and the functions
these attitudes serve?
Hypothesis la: There will be sex, age, education, religious ideology, religious
attendance, religious affiliation, media images, geographic residence and degree
of closeness with lesbians and gay men differences for heterosexuals’ attitudes
toward gay men and heterosexuals’ attitudes toward lesbians.
Hypothesis lb: Heterosexual men will have less favorable attitudes toward gay
men and toward lesbians than heterosexual women.
Hypothesis 1c: There will be a significant relationship between the number of
gay men and lesbians someone is close to and the defensive function.
Hypothesis Id: There will be gender group differences for the defensive
function.
Research Question 2: Are there sex differences in opinions about gay and lesbian civil
rights?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
Hypothesis 2: Heterosexual men will oppose civil rights for gay men and
lesbians more than will heterosexual women.
Research Question 3: Do attitude functions predict attitudes toward lesbians and
attitudes toward gay men?
Hypothesis 3a: The defensive function will predict less favorable attitudes
toward lesbians and toward gay men.
Hypothesis 3b: The social-expressive function will predict less favorable
attitudes toward lesbians and toward gay men.
Hypothesis 3c: The value-expressive function will predict less favorable
attitudes toward lesbians and toward gay men.
Hypothesis 3d: The experiential-schematic function will predict more favorable
attitudes toward lesbians and toward gay men.
Research Question 4: Are there sex and item order differences in attitudes toward
lesbians, attitudes toward gay men and attitude functions?
Hypothesis 4a: Heterosexual men’s attitudes toward lesbians will be more
unfavorable when questions about gay men are asked first.
Hypothesis 4b: Heterosexual men will exhibit an increased defensive function
when questions about gay men are asked before questions about lesbians.
Hypothesis 4c: Assessing attitudes about gay men first will make the defensive
function more accessible for male respondents, resulting in more unfavorable
attitudes toward gay men.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
CHAPTER III
Method
The goal of the present study was to examine the relationship between
demographic and social differences, attitudes toward gay men, attitudes toward
lesbians, opinions about civil rights for gay men and lesbians and the functions these
attitudes serve. Determining if there was a relationship between attitude functions and
attitudes was also a goal of this study. Another goal was to examine the relationship
between gender and the order attitudes and attitude functions were assessed. This
chapter describes the methodology used to examine these relationships.
Participants
Qualified participants were self-identified heterosexual adults residing in the
United States. A total of 411 qualified participants (202 men and 209 women)
participated in this study. Form A (lesbian attitude scales first and gay male attitude
scales second) was completed by 100 men and 106 women. Form B (gay male attitude
scales first and lesbian attitude scales second) was completed by 102 men and 103
women. The ages of participants ranged from 18 to 84, with a mean age of 44
(SD= 13.20). The majority of respondents, 56%, reported being married, 20% were
single, 16% were divorced, 3% were partnered, 3% were widowed and 2% were
separated. Eighty-seven percent of the respondents self-identified as White, 2%
Hispanic, 2% Asian, 3% African-American, 3% mixed racial background and 1% were
another race.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
Twenty-nine percent of the respondents never attended a religious ceremony
over the past year, 36% attended once to a few times, 10% attended 1 - 3 times per
month and 22% attended weekly or more often. Twenty-five percent were Catholic,
19% Protestant, 13% Baptist, 11% Bom-again Christian, 5% Jewish, 2% Agnostic, 1%
Buddhist, 4% spiritual with no specific affiliation and 18% described themselves as
“Other.”
Thirteen percent reported 2002 household incomes of $0 to $34,999, 65% had
household incomes between $35,000 and $99,999 and 15% reported household incomes
of $100,000 or more. Fifty percent of the sample did not hold a college degree, 28%
held an associate’s or college degree, 8% completed some graduate school, but did not
have a degree and 15% had a graduate school degree. Thirty-six percent of the
respondents were from a suburban area next to a city, 29% lived in an urban or city area
and 35% lived in a small town or rural area.
Table 1 compares the demographic characteristics of the respondents in this
study to those of the 2000 United States Census (U.S. Census Bureau, 2002). This
study’s male and female percentages are identical to the percentages in the United
States’ overall population in 2000. The racial make-up of this study’s sample is
comprised of 87% White respondents versus 75% Whites in the entire population.
African Americans and Hispanics are underrepresented in this study by a margin of 9%
and 10% respectively. The 411 adults who participated in this study were more
educated and had higher household incomes than those of the general population. The
four-year difference in household income between the Census data (the most recent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
88
household income information is for the year 1999) and these results from 2003 may
partially explain why this study’s sample reported higher incomes compared to the
general population. The marital statuses in this study are comparable to those in the
general population, except for those in this sample who reported being divorced did so
7% more than those in the overall population.
Table 1
Profile o f General Demographic Characteristics of the 2000 United States Census and Current Study
% (Current study) % (2000 census)
Sex
Male
Female
49.1
50.9
49.1
50.9
Race
White 86.9
African American 3.2
Hispanic 2.4
Native American or Alaskan Native .2
Asian or Pacific Islander 2.4
Mixed racial background 2.7
Two or more races
Other race 1.0
75.1
12.3
12.5*
.9
3.7
2.4
5.5
Education
Less than high school .5
Completed some high school 1.0
High school graduate or equivalent 14.8
Completed some college, no degree 33.3
Associate’s degree 10.0
Bachelor’s degree 17.5
Completed some graduate school,
No degree 7.5
Graduate or professional degree 15.3
7.5
12.1
28.6
21.0
6.3
15.5
8.9
Geographic Residence
In an urban or city area 29.2
In a suburban area next to a city 36.0
In a small town or rural area 34.8
Household Income*
Less than $15,000
$15,000 to $24,999
$25,000 to $34,999
$35,000 to $49,000
10.9
6.3
12.7
19.7
15.8
12.8
12.8
16.5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
Table 1—Continued
$50,000 to $74,999
$75,000 to $99,999
$100,000 to $149,999
$150,000 to $199,999
$200,000 or more
23.4
12.4
10.4
1.7
2.5
19.5
10.2
7.7
2.2
2.4
Marital Status
Single, never married 20.0
56.0
16.3
1.5
3.4
2.9
27.1
54.4
9.7
2.2
6.6
Married
Divorced
Separated
Widowed
Partnered
Note: *total population adds up to over 100% because the Hispanic or Latino population is of any race.
**Household income for the 2000 U.S. Census is for 1999.
Fifty-eight percent and 63% of the respondents were close to no gay men and no
lesbians respectively. Eighteen percent were close to one gay man and one lesbian, and
23% and 19% were close to two or more gay men and two or more lesbians
respectively. Nineteen percent and 31% of the respondents were acquaintances with no
gay men and no lesbians respectively, 29% and 22% were acquaintances with one gay
m an and one lesbian respectively and 52% and 46% of respondents were acquaintances
with two or more gay men and two or more lesbians respectively.
The data were collected by using two forms of a questionnaire (see Appendix A)
that altered the order of attitude questions about lesbians and about gay men depending
upon random assignment. The first section of each questionnaire (Form A and Form B)
included demographic and social item assessments: gender, race, education, geographic
residence, household residence, income, age, relationship status, number of children,
Instrument
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
sexual orientation, religious attendance, religious ideology, religious affiliation, contact
with lesbians and gay men, quality of experiences with gay men and lesbians, hours of
television viewing, assessment of media images of gay men and lesbians and ranking of
six factors that influence attitudes toward gay men and toward lesbians.
In Form A, the second section assessed attitudes toward lesbians (ATL) and
attitude functions for lesbians (AFI-L). The third section assessed attitudes toward gay
men (ATG) and attitude functions for gay men (AFI-G). In the final section of Form A,
respondents were asked to indicate their agreement or disagreement with the following
questions: (a) “Same-sex couples should be able to legally wed.” (b) “Gay men and
lesbians should be able to adopt children.” (c) “Lesbians and gay men should be legally
protected from prejudice and job discrimination in employment.” and (d) “Lesbians and
gay men should be able to serve openly in the military.”
In Form B, ATG and AFI-G were assessed first and the order in which
respondents were asked to assess attitudes was switched. The ATL and AFI-L were
assessed second. In the final section, the same four public policy questions were asked
that were asked in Form A.
Attitudes toward gay men and lesbians scales. The 10-item Attitudes Toward
Lesbians (ATL; Herek, 1984b) and Attitudes Toward Gay Men (ATG; Herek, 1984b)
scales were used to measure attitudes toward these two attitude targets. For each
statement, respondents were provided with five response alternatives (strongly disagree,
somewhat disagree, neither agree nor disagree, somewhat agree, strongly agree) on a
five-point Likert-type scale. Item responses were reversed as necessary and summed to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
yield a sub-scale score that could range from 10 (extremely favorable attitudes) to 50
(extremely unfavorable attitudes). Higher scores indicate more unfavorable attitudes
toward gay men and lesbians.
The ATLG, ATG and ATL scales were developed through a series of factor-
analytic studies by Herek (1984b). Internal consistency for the ATLG, ATG and ATL
was found to be satisfactory. The alpha coefficient for the ATLG was .90, ATG was
.89 and ATL was .77. Internal reliability for this study’s ATG scale had a coefficient
alpha of .94 and the ATL scale produced a coefficient alpha of .94. These coefficient
alphas are comparable and are slightly higher than those obtained in Herek’s (1984b)
study, suggesting the ATG and ATL were highly reliable in this study.
Attitude function scales—revised. The 10-item Attitudes Functions Inventory
(API; Herek, 1987) was used to measure four attitude functions for attitudes toward gay
men and lesbians. An example of a question used to assess the experiential-schematic
function is, “My opinions about lesbians mainly are based on whether or not someone I
care about is gay or lesbian.” A social-expressive question is, “My opinions about gay
men mainly are based on my perceptions of how the people I care about have responded
to gay people as a group.” A question asked to assess the defensive function is “My
opinions about gay men mainly are based on my personal feelings of discomfort or
revulsion at homosexuality.” And, one of the questions used to assess the value-
expressive function is “My opinions about gay men mainly are based on my moral
beliefs about how things should be.”
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
To test the reliability and validity of the API, Herek (1987) created three
versions of the measure targeting attitude functions of persons belonging to three
stigmatized groups: people with AIDS, people with cancer and mentally ill individuals.
Factor analysis for the four functions accounted for 70% of the total variance and the
items clustered as expected. Factor analyses demonstrated that the AFI items are
tapping four independent constructs, namely four differing psychological motivations
behind attitudes.
The AFI has been shown to have acceptable levels of reliability and validity
(e.g., Herek, 1987; Shavitt, 1990; Maio & Olson, 1994, 1995). Herek (1987) found
alpha coefficients for three different attitude objects were: .81 (experiential-schematic);
.75 (social-expressive); .80 (defensive) and .87 (value-expressive). These relatively
high alpha coefficients indicate a significant degree of consistency in functions across
three attitude domains (AIDS, mental health and cancer). This stability is likely due in
part to the fact that all three of the attitude objects assessed involved persons belonging
to stigmatized groups.
For this study, the coefficient alpha for the experiential-schematic function was
.87, for the social-expressive function the coefficient alpha was .87, for the defensive
function the coefficient alpha was .88 and for the value-expressive function the
coefficient alpha was .63. These coefficient alphas are acceptable and comparable to
those found by Herek (1987), except for the value-expressive function coefficient alpha,
which appears to be lower than in Herek’s study. A low coefficient alpha implies that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
the items measuring the factor are not highly correlated, meaning the sub-scale may not
be measuring what it purports to measure.
This study used two variations of the AFI. One version assessed attitude
functions for attitudes toward lesbians. The second version assessed attitude functions
for attitudes toward gay men. For each statement, respondents were provided with five
response alternatives (strongly disagree, somewhat disagree, neither agree nor disagree,
somewhat agree, strongly agree).
The AFI can be scored by assigning respondents to groups on the basis of their
dominant attitude functions or by treating each subscale as a continuous variable. The
latter approach was used for this study. Scores for the four attitude function subscales
were obtained by summing the items within the individual subscales and dividing the
total by the number of items within that subscale. Higher scores indicate endorsement
of a particular attitude function. For the purposes of this study, the defensive, social-
expressive, value-expressive and experiential-schematic functions were continuous
variables.
Procedure
The data were self-reported and collected by means of questionnaires completed
through the Internet between June 26 and June 30 of 2003 with one reminder sent on
June 29. Online invitations to participate in this study were sent to 2,575 panelists. The
response rate for this study was 16% and it took respondents on average 14 minutes to
complete. A total of 74 participants began the survey but did not complete it.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
Those invited to participate in the study were randomly drawn from
approximately three million people who had voluntarily elected to participate in Internet
research, provided Harris Interactive with their e-mail addresses and opted into the
Harris Poll Online Panel. Participants were informed of the purpose of this study and
were told in writing that participation in completing the online survey was voluntary.
They were informed that online submission of the questionnaire constituted consent to
participate in the study and were assured of confidentiality. Qualified participants who
chose to complete the survey received 100 points toward their HI Stakes account. HI
Stakes accounts are an incentive account that is administered and maintained by Harris
Interactive. The account allows panelists to redeem points for a variety of consumer
products. Non-qualified participants were those who reported a sexual orientation other
than heterosexual, were less than 18 years of age or resided outside of the United States.
Non-qualified participants received 30 points. Anyone invited to participate in the
study was given the option to participate in a sweepstakes administered by Harris
Interactive.
Participants living outside of the United States, those who indicated they were a
sexual orientation other than heterosexual and those under the age of 18 were
administered a shortened survey. These 68 respondents’ responses were not used in this
study.
Qualifying participants were randomly assigned to receive one of two forms.
Form A requested information about their attitudes and attitude functions toward
lesbians first followed by the same questions about gay men. Form B asked questions
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
about their attitudes and attitude functions toward gay men before their attitudes and
attitude functions toward lesbians.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
CHAPTER IV
Results
This chapter presents the results of the study. Descriptive analyses of the
study’s variables, including means, standard deviations and intercorrelations are
provided first. Then, results from the MANOVAs and stepwise multiple regression
analyses for the four research questions are provided and summarized. Sections
include: (a) differences between attitudes and attitude functions by sex, residence,
closeness to gay men and lesbians, religion and media images, (b) sex differences in
opinions about public policy issues, (c) attitude prediction by attitude function, and (d)
sex differences in attitudes and attitude functions by item order.
Intercorrelations
The means, standard deviations and correlations of the primary variables are
presented in Table 2. Sex was positively correlated with closeness to gay men (r=. 14,
pc.Ol), indicating that heterosexual women in this study reported being closer to more
gay men than their heterosexual male counterparts. Sex was also significantly
correlated with the Attitudes Toward Lesbians (ATL; r=-.14, pc.Ol) and Attitudes
Toward Gay Men scales (ATG; r=-,18,p<.01), which suggests women had lower sexual
prejudice scores than did men, meaning they held more favorable attitudes toward
lesbians and toward gay men.
Further, sex was positively correlated with the experiential-schematic function
subscale for gay men (AFI-G; r=.10, p<.05) and the defensive function for gay men
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
subscale (AFI-G; r=-.15,p<.01). These correlations suggest that heterosexual women
scored higher than men on the experiential-schematic function meaning women
reported that they hold their attitudes toward gay men based on their experiences with
and knowledge of gay men more so than men do. Heterosexual men scored higher than
women on the defensive function when answering questions about the motivation
behind their attitudes toward gay men. This suggests that heterosexual men may be
more motivated than women are to hold their attitudes toward gay men due to personal
feelings of discomfort or revulsion by homosexuality.
There was a positive correlation between closeness to gay men and closeness to
lesbians (r=.44, pc.Ol). For the purposes of this study, closeness was defined as the
number of gay men or lesbians the respondent speaks to on a regular basis and is
emotionally tied to in some way. There were negative correlations between closeness to
gay men and attitudes toward lesbians (ATL; r=-.26,p<.01) and attitudes toward gay
men (ATG; r=-.27, pc.Ol). Further, closeness to lesbians was negatively correlated
with attitudes toward lesbians (ATL; r=-.20, pc.Ol) and attitudes toward gay men
(ATG; r=-.18,/?<.01). These correlations suggest that as respondents were closer to
more lesbians and gay men, attitudes toward lesbians and gay men tended to improve.
Further, there were significant negative correlations between the number of gay men
(AFI-L; r=-.21, pc.Ol; AFI-G; r=-.22, pc.Ol) and number of lesbians (AFI-L; r=-.20,
pc.01; AFI-G; r=-.15,jx.01) respondents were close to and the defensive function for
lesbians and for gay men. Additionally, there was a significant correlation between
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 2
Zero-order Pearson Product Correlations for Demographic and Measured Variables
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 ' 17
1. Sex - -.05 -.13* .14** -.01 -.18** -.04 -.14** -.18** .07 -.04 -.04 .05 .10* -.05 -.15** -.03
2. Age -.06 -.11 .05 -.10 .08 .09 -.03 .02 .04 -.03 -.06 -.06 .08 -.02
3. Income -.12* -.07 .27** -.02 .05 .04 -.01 .04 .03 .03 -.00 .03 .06 .07
4. Closeness to
Gay Men
_
.44* .09 .09 -.26** -.27** .10 .03 -.21** .04 .20** .04 -.22** -.01
5. Closeness to
Lesbians -.01 -.01 -.20** -.18** .08 .05 -.20** .03 .04 .02 -.15** .02
6. Education — —
.08 -.06 -.08 .04 .11* -.07 .09 .04 .12* -.07 .13*
7. Item Order — .12* .11* .08 .12* .12* .10* .03 .08 .08 .05
8. ATL — .94* -.22** .01 .50** .05 -.22** .04 .52** .08
9. ATG — -.21* .03 .50** .05 -.19** .05 .54** .09
10. Experiential
Schematic
Function-L
. .
.49** .18** .07 .45** 16** .09
11. Social-Exp.
Function-L .37** .18** .39** .78** 33** .22**
12. Defensive
Function-L
_
.19** .17** .35** .81** .18**
13. Value-Exp.
Function-L
—
.08 .16** .17** .75**
14. Experiential
Schematic
Function-G
- .49** 21** .09
15. Social-Exp.
Function-G
.40** .25**
16. Defensive
Function-G
„ .23**
17. Value-Exp.
Function-G
Note. All scores are scaled scores. 7 = Item Order: ATL and AFI-L Questions Before ATG and AFI-G Questions and Vice Versa, 8 = Attitudes Toward
Lesbians (ATL), 9 = Attitudes Toward Gay Men (ATG), 10-17 = Four Attitude Functions for Lesbians and Gay Men. *. Correlation is significant at the
.05 level; **. Correlation is significant at the .01 level.
' O
oo
99
closeness to gay men and the experiential-schematic function for gay men (AFI-G;
r=.20, p<01). There were significant positive relationships between education and the
social-expressive function for lesbians (AFI-L; r=.ll,/?<.05) and for gay men (AFI-G;
r=.12,p<.05) as well as the value-expressive function for gay men only (AFI-G; r=. 13,
j?<.05). This could indicate that more highly educated people tend to incorporate their
social group’s opinions into their own evaluation of gay men and lesbians.
There was a significant and strong correlation between attitudes toward lesbians and
attitudes toward gay men (r=.94, p<,05). Attitudes toward lesbians were significantly
associated with the experiential-schematic function (AFI-L; r=-,22, AFI-G; r=-.22,
jk.01) and defensive function (AFI-L; r=.50, pc.Ol; AFI-G; r=.52, pc.Ol). This
suggests that respondents who had higher scores on the experiential-schematic attitude
function sub-scales had more favorable attitudes toward lesbians and those with higher
scores on the defensive function had less favorable attitudes toward lesbians. Similar
correlations were found for attitudes toward gay men and the experiential-schematic
(AFI-L; r=-.21,p<.01; AFI-G; r=-.19, pK.Ol) and defensive functions (AFI-L; r=.50,
pK.Ol; AFI-G; r=.54,p<.01).
Research Question 1: Are There Demographic and Social Differences That Predict
Heterosexuals’ Attitudes Toward Lesbians, Attitudes Toward Gay Men and the
Functions These Attitudes Serve?
Hypothesis la: There will be sex, age, education, religious attendance, religious
ideology, religious affiliation, media images, geographic residence and degree of
closeness with lesbians and gay men differences for heterosexuals’ attitudes toward gay
men and heterosexuals’ attitudes toward lesbians. Table 3 presents mean ATL and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
ATG scores and standard deviations by respondent sex, residence, religious attendance,
religious ideology, closeness to gay men and closeness to lesbians. Due to the small
Table 3
Mean ATL and ATG Scores by Respondent Sex, Residence, Religious Attendance, Religious Ideology,
Religious Affiliation, Closeness to Gay Men and Closeness to Lesbians
Factor M
Scale
ATL
SD
ATG
M SD
Sex
Women («=209) 21.57 11.33 21.99 11.23
Men («=202) 24.65 11.04 26.08 11.80
Residence
Urban (n=120) 21.37 10.66 22.46 11.47
Suburban (n=148) 22.75 10.96 23.28 11.32
Rural (n=143) 24.87 11.91 26.04 12.03
Religious Attendance
Never («= 117) 19.15 10.19 20.26 11.48
Once to a few times (n=148) 21.08 10.15 22.08 10.45
1-3 times per month (n=41) 22.93 10.39 22.95 9.99
Weekly or more often (n=90) 31.02 10.92 31.81 10.87
Religious Ideology
Conservative («=180) 27.19 11.40 28.37 11.90
Liberal (n=81) 19.27 9.32 19.65 9.32
Other (n=44) 21.84 10.92 23.30 11.20
No affiliation («=58) 19.45 10.28 20.36 10.31
Not religious («=41) 18.80 10.78 19.22 11.38
Religious Affiliation
Catholic (n=77) 22.23 9.94 23.65 10.86
Jewish (n=16) 17.81 7.09 16.56 6.77
Born-again Christian (n=32) 34.53 9.88 34.75 9.86
Baptist (n=40) 27.08 12.11 28.73 12.87
Buddhist (n=3) 11.33 2.31 14.67 4.04
Protestant (n=57) 23.30 10.10 24.02 9.75
Agnostic (n=5) 19.20 8.70 21.80 9.52
Other Religion (n=55) 25.31 12.42 26.22 12.77
Spiritual w/no Specific
Religious Affiliation (n=13) 18.85 11.16 18.54 11.13
# of Close Gay Men
0 (n=208) 24.98 11.20 26.11 11.57
1 (n=66) 20.03 10.24 20.59 10.31
2 («=49) 16.10 7.01 16.61 6.74
3+ («=34) 16.24 8.09 16.74 8.73
# of Close Lesbians
0 («=228) 24.18 11.14 24.95 11.67
1 (n=63) 19.33 9.91 20.11 9.27
2 («=45) 17.29 8.47 18.51 9.73
3+ (n=23) 16.96 9.23 18.74 9.95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
number of respondents in the closeness to 3,4 and 5 or more categories for both
lesbians and gay men, these categories were combined. The new categories, being close
to 3 or more gay men and being close to 3 or more lesbians, had a sufficient number of
cases in order to produce more powerful results. The correlations for education, age
and daily television viewing duration and attitudes toward gay men and attitudes toward
lesbians were not significant. Therefore, no further analyses were conducted on these
three independent variables.
Conducting five separate one-way MANOVAs for the independent variables of
sex, residence (urban or city area, suburban area next to a city, small town or rural area),
religious attendance in the past year (never, once to a few times, 1-3 times per month,
weekly or more often), religious ideology (conservative, liberal, other, no religious
affiliation, not religious) and religious affiliation (Catholic, Jewish, Bom-again
Christian, Baptist, Buddhist, Protestant, Agnostic, Other religion, Spiritual with no
specific religious affiliation) tested hypotheses la and lb. To test hypothesis la, an
additional two separate two-way MANOVAs were conducted for the independent
variables of assessment of images of gay men and lesbians in the media and closeness
with gay men and lesbians for the measures of attitudes toward lesbians and attitudes
toward gay men. No significant differences were found in two of the
analyses—residence, assessment of media images of gay men and assessment of media
images of lesbians. For the statement, “Overall, the images of gay men that I have seen
in the media over the past two years have portrayed gay men in a Jight.” One
hundred ninety-eight respondents indicated they thought the images were positive or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
somewhat positive and 87 respondents indicated they thought gay male images were
portrayed in a negative or somewhat negative light. Respondents rated images of
lesbians in the media less favorably. One hundred sixty-five of the respondents rated
lesbian images as positive or somewhat positive, while 99 rated them negative or
somewhat negative.
Statistical significance was obtained for the MANOVAs that examined sex,
religious attendance, religious ideology, religious affiliation, closeness to gay men and
closeness to lesbians. To control the overall error rate, a Bonferroni correction was
calculated by using the error rate for each test, p<05, divided by the total number of
tests, in this case that was 5. For findings to be significant, the minimum requirement
of pc.Ol must be met. Table 4 contains the MANOVA results.
The results of the MANOVA revealed significant differences in religious service
attendance and attitudes toward lesbians and attitudes toward gay men, Wilks’
Lambda=.835, F(3, 6) = 7.99, p<.001. Additionally, post hoc Scheffe analyses revealed
a significant difference in attitudes toward gay men and attitudes toward lesbians
between those who attend religious services weekly or more often and those who
indicated they never attend, attend once to a few times each year and attend one to three
times per month. Meaning, those who attend religious services weekly or more tend to
hold more unfavorable attitudes toward gay men and toward lesbians than those who
attend religious services less frequently.
In order to gain a greater understanding of respondents’ religious practices, they
were asked to self-identify as religiously conservative, religiously liberal or to choose
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
from other descriptions of religious ideologies. They were also asked about their
religious identification in terms of specific religious denominations such as Catholic,
Jewish and Protestant.
Table 4
Multivariate Analysis of Variance for ATL and ATG by Sex, Residence,
Religious Attendance, Religious Ideology and Closeness to Gay Men and
Closeness to Lesbians
Source Wilks’ F df
P
Sex .962 7.99 2 .000
ATL 7.76 1 .006
ATG 13.01 1 .000
Sex .973 4.98 2 .007
Closeness to Gay Men 7.35 1 .007
Closeness to Lesbians .058 1 .810
Residence .978 2.24 4 .063
ATL 3.29 2 .038
ATG 3.56 2 .029
Religious Attendance .835 12.33 6 .000
ATL 25.05 3 .000
ATG 21.97 3 .000
Religious Ideology .875 5.59 10 .000
ATL 10.17 5 .000
ATG 11.01 5 .000
Religious Affiliation .815 3.82 16 .000
ATL 6.37 8 .000
ATG 6.06 8 .000
Closeness to Gay Men .953 2.75 6 .012
ATL 3.77 3 .011
ATG 5.20 3 .002
Closeness to Lesbians .976 1.38 6 .221
ATL 2.30 3 .077
ATG 1.51 3 .211
There were religious ideology group differences among mean scores of attitudes
toward gay men and attitudes toward lesbians, Wilks’ Lambda=.875, F(5, 10) = 5.59,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
jx.001. As hypothesized, post hoc Scheffe analyses suggested that respondents with
conservative religious affiliations had significantly higher levels of sexual prejudice
than respondents with liberal religious affiliations, no religious affiliation or non
religious respondents.
The results of the MANOVA also revealed significant differences in attitudes
toward gay men and attitudes toward lesbians by religious affiliation, Wilks’
Lambda=.815, F(8, 16) = 3.882, jx.OGl. Due to the small number of respondents who
self-identified as Buddhist (n=3) and Agnostic (n=5), these two groups were left out of
the analysis. Post hoc Scheffe analyses revealed significant differences between Born-
again Christian’s attitudes toward gay men and lesbians and those respondents who
indicated they were Catholic (p = .000; mean difference = -11.10), Jewish (p = .001;
mean difference = -18.19), Protestant (p = .005; mean difference = -10.73) and spiritual
with no specific religious affiliation (p = .012; mean difference = -16.21). This means
that those respondents who identified as Catholic, Jewish, Protestant and spiritual with
no specific religious affiliation expressed significantly more favorable attitudes toward
lesbians and toward gay men than those who identified as Born-again Christian.
Compared to Born-again Christians, Jewish respondents and respondents indicating
they were spiritual with no specific religious affiliation had the highest mean difference
in ATL and ATG scores, suggesting their attitudes were far more favorable than were
the attitudes of Born-again Christians’ attitudes toward lesbians and toward gay men.
Statistical significance was obtained for closeness to gay men and attitudes
toward lesbians and attitudes toward gay men, Wilks’ Lambda=.953, F(3, 6) = 2.75,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
p=.012. Being close to one or more gay men in comparison with being close to no gay
men was associated with more favorable attitudes toward lesbians and more favorable
attitudes toward gay men. Significant differences were not found in the number of
lesbians respondents were close to and their attitudes toward lesbians and attitudes
toward gay men, Wilks’ Lambda=.976, F(3, 6) = 1.38,p=.221.
Further, significance was obtained for sex and closeness to gay men, Wilks’
Lambda=.973, F( 1, 2) = 4.98,p=.007. This means that men were closer to fewer gay
men than were women.
Hypothesis 1c: There will be a significant relationship between the number of
gay men and lesbians someone is close to and the defensive function. To test for
differences in closeness and gender group differences for the defensive function and
eight attitude function sub-scales, hypotheses lc and Id respectively, two separate one
way MANOVAs were conducted. Again, due to the small number of respondents in the
closeness to 3, 4 and 5 or more categories for both lesbians and gay men, these
categories were collapsed into one category, being close to 3 or more gay men and
being close to 3 or more lesbians. The results of the MANOVA for closeness and the
defensive function (see Table 5) revealed there were nearly significant differences
between the number of gay men respondents were close to and the defensive function
for gay men, Wilks’ Lambda=.966, F(3, 6) = 2.00, /?=.06. Pairwise comparisons
revealed that there were significant differences for the defensive function for gay men
between respondents who were close to no gay men and those who were close to two or
more gay men. Significance was not obtained for closeness to gay men and the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
defensive function for lesbians. These results suggest that being close to no gay men is
associated with an increased defensive function. No significant differences were found
in the closeness to lesbians analyses, Wilks’ Lambda=.986, F(3, 6) = .$16, p = .558.
Table 5
Multivariate Analysis of Variance for the Defensive Function By Sex and Closeness to
Gay Men and Closeness to Lesbi a n s _ ^ ____________
Source Wilks’ F
df p
Closeness to Gay Men .966 2.00 6 .064
Defensive-L 1.37 3 .252
Defensive-G 3.43 3 .017
Closeness to Lesbians .986 .816 6 .558
Defensive-L 1.37 3 .251
Defensive-G .85 3 .466
Sex .932 3.67 8 .000
Experiential-Schematic-L 1.84 1 .176
Social-Expressive-L .61 1 .436
Defensive-L .64 1 .424
V alue-Expressive-L .99 1 .320
Experiential-Schematic-G 4.23 1 .040
Social-Expressive-G .83 1 .362
Defensive-G 9.07 1 .003
Value-Expressive-G .26 1 .611
Hypothesis Id: there will be gender group differences for the defensive function.
The results of the MANOVA for the eight attitude function subscales revealed there
were significant sex differences, Wilks’ Lambda=.932, F{ 1, 8) = 3.66,p<.001. In
follow-up univariate analyses, only the experiential-schematic and defensive functions
for attitudes toward gay men were significant. These results suggest that when
heterosexual women report their motivation for holding their attitudes toward gay men,
they attribute their experiences with and knowledge of gay men significantly more than
do heterosexual men. In other words, heterosexual women endorse the experiential-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
schematic function for attitudes toward gay men significantly more than do
heterosexual men (M= 2.58 for women, M=2.38 for men). Further, the results suggest
the opposite is true for the defensive function (M=2.35 for men, M -2.00 for women).
Heterosexual men tend to attribute their revulsion of and discomfort with
homosexuality as a reason for holding their attitudes toward gay men significantly more
than heterosexual women do.
Research Question 2: Are There Sex Differences in Opinions About Gay and Lesbian
Civil Rights?
When asked about public policy questions related to gay and lesbian issues (see
Table 6), the majority of respondents (48%) agreed with the statement, “Same-sex
couples should be able to legally wed.” Fifty-seven percent agreed that gay men and
lesbians should be allowed to adopt children and 36% opposed this statement. Eighty
percent believed that lesbians and gay men should be legally protected from prejudice
and discrimination in employment. Only 17% of the respondents disagreed with this
statement. Sixty-nine percent of those asked agreed that lesbians and gay men should
be able to serve openly in the military, with 26% against lesbians and gay men serving
openly in the military.
Hypothesis 2: Heterosexual men will oppose civil rights for gay men and
lesbians more than will heterosexual women. To test for sex differences in respondents’
agreement or disagreement with four public policy issues related to gay men and
lesbians, a one-way MANOVA was conducted. As hypothesized and consistent with
past results, significant sex differences were found for each of the four questions,
Wilks’ Lambda=.941, F (l, 4) = 5.24, pc.OOl. Table 7 shows that men are less likely
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
Table 6
Comparison of Past Responses to Four Public Policy Questions and Respondent Sex
Breakdown for this Study
Same-sex couples should be able to legally wed.
1998 2000 2003
Men Women
Agree 33% 39% 48% n=87 »=112
Disagree 55% 45% n -107 «=77
Gay men and lesbians should be allowed to adopt children.
1994 2000 2003
Men Women
Agree 29% 46% 57% n=99 n=137
Disagree 47% 36% n=89 n=57
Lesbians and gay men should be legally protected from
prejudice and discrimination in employment.
1977 1996 2000 2003
Men Women
Agree 56% 66% 76% 80% n=151 n=177
Disagree 17% n=45 «=24
Lesbians and gay men should be able to serve openly in the military.
2000 2003
Men Women
Agree 56% 69% n=118 n=166
Disagree 26% n— 76 n -32
Note: 1977,1996 polls are cited in Rogers, 1998; 1994, 1998 polls cited in The Henry J. Kaiser
Family Foundation, 2001); 2000 study was conducted by The Henry J. Kaiser Family
Foundation, 2001.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
than women to agree that lesbians and gay men should have legal rights and protection
for marriage, adoption, job prejudice and discrimination and serving openly in the
military.
Table 7
Multivariate Analysis of Variance for Four Public Policy Issues by Sex _ _ _
Source Wilks’ F df p
~Sex _ - _ _ _ _
Marriage 9.22 1 .003
Adoption 10.79 1 .001
Job Discrimination 27.92 1 .005
Military 18.82 1 .000
Research Question 3: Do Attitude Functions Predict Attitudes Toward Lesbians and
Attitudes Toward Gay Men?
To determine what functions best predict attitudes toward lesbians and attitudes
toward gay men, stepwise multiple regression analyses were conducted for each of the
attitude function subscales. Table 8 illustrates the variance explained by all of the
attitude functions. Tables 9 through 12 show the contributions of each attitude function
on heterosexual men’s and heterosexual women’s attitudes toward lesbians and attitudes
toward gay men.
Table 9 summarizes the results of the stepwise multiple regression, which
revealed that of the eight attitude functions, the best predictor for heterosexual men’s
attitudes toward lesbians was the defensive function for gay men, which explained
22.9% of the variance, F change (1, 200) = 59.27, p = .000. The experiential-
schematic function for lesbians explained an additional 11.2% of the variance, F change
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
(1, 199) = 33.78, p = .000. Further, the defensive function for lesbians explained 2.1%
of the variance in addition to the 34% already explained by the other two functions, F
change (1, 198) = 6.50, p = .012. These results suggest that heterosexual men report
that they form their attitudes toward lesbians primarily based on their discomfort
Table 8
Variance Explained by Attitude Functions Predicting Men’s and
Women’s ATL and ATG Scores
Variables
Men
ATL ATG
Women
ATL ATG
Experiential-S chematic-L 11.2% 10.2%
Social-Expressive-L 1.2%
Defensive-L 2.1% 2.9% 2.9%
V alue-Expressive-L
Experiential-S chematic-G 11.3% 9.0%
Social-Expressive-G
Defensive-G 22.9% 24.5% 30.3% 32.6%
V alue-Expressive-G
of and revulsion by gay people and to a far lesser degree by their experiences with and
closeness to gay people.
Table 9
Summary o f Stepwise Regression Analysis for M en’ s Attitudes Toward Lesbians
Variable R R Change F Change Beta
P
Defensive-G .229 59.27 .34 .000
Experiential-
S chematic-L .341 .112 33.78 -.35 .000
Defensive-L .361 .021 6.50 .27 .012
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I l l
Table 10 shows the results of the stepwise multiple regression for men’s
attitudes toward gay men. The results revealed that the defensive function for gay men
was the best predictor of attitudes toward gay men, accounting for 24.5% of the
variance, F change (1, 200) = 64.87, p = .000. The experiential-schematic function for
gay men accounted for an additional 10.2% of the variance, F change (1,199) = 31.15,
Table 10
Summary o f Stepwise Regression Analysis for M en’ s Attitudes Toward Gay Men
Variable R R Change F Change Beta p
Defensive-G .245 64.87 .57 .000
Experiential-
Schematic-L .347 .102 31.15 -.33 .000
p = .000. These results suggest that heterosexual men’s attitudes toward gay men are
primarily motivated by their revulsion of and discomfort with homosexuality. These
results also suggest that experiences with and closeness to gay men and lesbians also
play a significant role in how heterosexual men report why they hold their attitudes
toward gay men.
The results of the stepwise multiple regression for women’s attitudes toward
lesbians are shown in Table 11. The results revealed that the defensive function for gay
men was the best predictor of attitudes toward lesbians, accounting for 30.3% of the
variance, F change (1, 200) = 90.09, p = .000. The experiential-schematic function for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
gay men contributed an additional 11.3% to the variance explained, F change (1,199) =
39.86, p = .000. In addition to the 41.6% of the variance already explained, the
defensive function for lesbians explained an additional 2.9% of the variance, F change
(1, 198) = 10.84, p = .001.
Table 11
Summary o f Stepwise Regression Analysis fo r Women’ s Attitudes Toward Lesbians
Variable R2 R Change F Change Beta
P
Defensive-G .303 90.09 .38 .000
Experiential-Schematic-G .416 .113 39.86 -.34 .000
Defensive-L .446 .029 10.84 .28 .001
The results of the stepwise multiple regression for women’s attitudes toward gay
men are shown in Table 12. The results revealed that the defensive function for gay
men was the best predictor of women’s attitudes toward gay men, accounting for 32.6%
of the variance, F change (1, 200) = 100.20, p = .000. The experiential-schematic
function for gay men contributed an additional 9.0% to the variance explained, F
change (1, 199) = 3 1.64, p = .000. In addition to the 41.6% of the variance already
explained, the defensive function for lesbians explained an additional 2.9% of the
variance, F change (1, 198) = 10.85, p = .001. Finally, the social-expressive function
for lesbians accounted for an additional 1.2% of the variance explained, F change (1,
197) = 4.60, p — .033.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
Hypothesis 3a: The defensive function will predict less favorable attitudes
toward lesbians and toward gay men. The positive Beta values for all four multiple
regression analyses support this hypothesis.
Hypothesis 3b: The social-expressive function will predict less favorable
attitudes toward lesbians and toward gay men. The valence of the Beta in the one
condition when the social-expressive function was predictive of attitudes (see Table 12),
Table 12
Variable R Change F Change Beta
P
Defensive-G .326 100.20 .38 .000
Experiential-
Schematic-G .416 .090 31.64 -.26 .000
Defensive-L .445 .029 10.85 .31 .001
Social-Expressive-L .457 .012 4.60 -.12 .033
for women’s attitudes toward gay men, did not support this hypothesis. Caution should
be used in interpreting this result due to the low absolute value of the Beta.
Hypothesis 3c: The value-expressive function will predict less favorable
attitudes toward lesbians and toward gay men.
Hypothesis 3d: The experiential-schematic function will predict more favorable
attitudes toward lesbians and toward gay men. The negative Beta values for all four
multiple regression analyses support this hypothesis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
Research Question 4: Are There Sex and Item Order Differences in Attitudes Toward
Lesbians, Attitudes Toward Gay Men and Attitude Functions?
Table 13 shows the means and standard deviations for the ATL and ATG by sex
and item order. Heterosexual men had the highest sexual prejudice scores toward both
lesbians and gay men when questions about gay men were asked first. ATG scores
were the highest of the two. When questions about lesbians were asked first,
heterosexual women scored the lowest on the ATL and second lowest on the ATG
scale.
Table 13
Mean ATL and ATG Scores by Respondent Sex and Item Order
Scale
ATL first
(«=206)
Women (n=106) Men («= 100)
ATG first
(n=205)
Women («=1Q3) Men (r=1Q2)
ATL
M 20.79 22.85 22.38 26.41
SD 11.09 10.30 11.56 11.50
ATG
M 21.35 24.16 22.64 27.97
SD 11.24 11.51 11.25 11.84
Table 14 shows mean scores for attitude functions and their standard deviations
by respondent sex and item order. Overall, respondents scored highest on the value-
expressive function (scores ranged from 3.19 to 3.55 out of the highest possible score of
5.00), followed by the experiential-schematic function (scores ranged from 2.29 to
2.62), then the defensive function (scores ranged from 1.95 to 2.49) and the lowest
scores for the social-expressive function (scores ranged from 1.94 to 2.30). These
scores appear to be consistent with respondents’ ranking of factors influencing their
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
attitudes toward gay men and lesbians. The factor of personal values was ranked
highest, which is related to the value-expressive function. Personal values were
followed by experiences with gay men and lesbians, which is related to the experiential-
schematic function.
Hypothesis 4a: Heterosexual men’s attitudes toward lesbians will be more
unfavorable when questions about gay men are asked first. Hypothesis 4b:
Table 14
Mean AFI-L and AFI-G Scores by Respondent Sex and Item Order
AFI-L first AFI-G first
Women Men Women Men
0=106) 0=100) (n=103) 0=102)
Experiential-Schematic-L
M 2.35 2.29 2.56 2.37
SD .87 .87 1.03 .99
Social-Expressive-L
M 1.98 2.03 2.19 2.30
SD .97 .98 1.10 1.06
Defensive-L
M 1.96 2.02 2.20 2.32
SD 1.08 1.12 1.19 1.22
V alue-Expressive-L
M 3.38 3.19 3.50 3.49
SD 1.04 .96 1.15 .95
Experiential-Schematic-G
M 2.55 2.36 2.62 2.39
SD 1.06 1.01 .99 .96
Social-Expressive-G
M 1.94 2.07 2.14 2.20
SD 1.00 1.03 1.05 .95
Defensive-G
M 1.95 2.22 2.06 2.49
SD 1.09 1.26 1.14 1.19
V alue-Expressive-G
M 3.34 3.27 3.33 3.55
SD 1.13 1.00 1.16 .89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
Heterosexual men will exhibit an increased defensive function when questions
about gay men are asked before questions about lesbians. Hypothesis 4c: Assessing
attitudes about gay men first will make the defensive function more accessible for male
respondents, resulting in more unfavorable attitudes toward gay men. To examine sex
and item order differences, a two-way MANOVA was conducted for measures of
attitudes toward lesbians, attitudes toward gay men and attitude functions. Table 15
shows that no significant item order differences were obtained in any of the analyses
although item order approached significance for ATL and ATG scores, Wilks’
Lam bda-987, F(l, 2) = 2.75, p = .065. No further analyses were conducted.
Table 15
Multivariate Analysis of Variance for ATL, ATG, AFI-L, AFI-G by Item Order
and Sex
Source Wilks’ F d f p
IteinOder-" '
ATL 5.50 1 .020
ATG 5.09 1 .025
Item Order x Sex .997 .71 2 .493
ATL .81 1 .473
ATG 1.24 1 .975
Item Order x Sex .968 1.68 8 .102
Experiential-Schematic-L 2.43 1 .120
Social-Expressive-L 5.99 1 .015
Defensive-L 5.84 1 .016
Value-Expressive-L 4.05 1 .045
Experiential-Schematic-G .27 1 .602
Social-Expressive-G 2.64 1 .105
Defensive-G 2.73 1 .100
Value-Expressive-G ___ _________ 1.09 1 .298
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
The results of the MANOVA for the ATL and ATG are inconsistent with the
findings from a study conducted by Herek (2000). He found that when lesbian items
were presented first, heterosexual men’s ATL scores were significantly lower than
when items about gay men were presented first. Consistent with this study, Herek
(2000) found no other significant differences in ATL and ATG scores by order and sex.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
CHAPTER V
Discussion
This study sought to examine the relationship of demographic and social
differences, attitudes toward gay men, attitudes toward lesbians, opinions about civil
rights for gay men and lesbians and the functions these attitudes serve. Determining if
there was a relationship between attitude functions and attitudes was also a primary goal
of this study. A secondary goal was to examine the relationship between gender and the
order attitudes and attitude functions were assessed.
Many factors contributed to the strength of this study’s generalizability, such as
the sample was relatively representative of the United States’ population. Most studies
examining attitudes toward gay men, attitudes toward lesbians or attitudes toward
homosexuals as one target group have relied on paper and pencil surveys with
geographically anchored college student populations.
This is the first study of its kind to use an online survey method with a
probability sample. Further, this is the first study asking questions about closeness to
gay men separate and apart from closeness to lesbians. This is also the first study
exploring the influence of asking questions in a different order on attitude functions.
This chapter discusses the findings of this study in the context of previous
research and theories about attitudes and gender. The findings’ implications, limitations
of the study as well as suggestions for future research are also discussed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
Discussion of Results
Hypothesis la: There will be sex, age, education, religious attendance, religious
ideology, religious affiliation, media images, geographic residence and degree of
closeness with lesbians and gay men differences for heterosexuals’ attitudes toward gay
men and heterosexuals ’ attitudes toward lesbians. Consistent with previous research,
significant differences in sex, religious attendance, religious ideology, religious
affiliation and closeness to gay men for attitudes toward gay men and attitudes toward
lesbians were found.
Surprisingly, no significant differences were found for age, education,
geographic residence, closeness to lesbians and media images. Most studies have found
that as age decreases, as respondents are more educated, close to two or more gay men
or lesbians and are from urban areas, attitudes toward gay men and toward lesbians
become more favorable. The three geographic resident descriptors were almost evenly
represented in this sample, the education of respondents closely resembled the United
States population and there were a wide range of ages (18 to 84) represented in this
study’s sample. Therefore, sample biases for age, education and residence are not
plausible explanations for why there were not significant group differences in attitudes
toward gay men and lesbians. Geographic residence approached significance.
One possible explanation for these findings factors in the influence the
Lawrence v. Texas decision had on respondents. The decision dominated local and
national media beginning the day panelists were invited to participate in this study. It is
possible that because the Supreme Court justices of the United States made a decision in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
favor of gay and lesbian rights, respondents may have taken the credibility of the
source, being the Supreme Court justices, and revised their immediate feelings and
thoughts about gay men and lesbians, regardless of their education, age and residence.
Overall, average ATL scores (M=24.65, 519=11.04) and ATG scores (M=26.08,
SD=l 1.80) were neither favorable nor unfavorable. The average response to attitude
scale items was “neither agree nor disagree” suggesting that prevailing attitudes toward
gay men and toward lesbians were relatively neutral to somewhat favorable. Most
studies have found attitudes toward gay men and lesbians to be far more unfavorable.
One explanation for this improvement in attitudes is the growing acceptance of gay men
and lesbians in American culture over the past few decades. An alternative explanation
is that sodomy laws being struck down only temporarily improved respondents’
attitudes toward gay men and lesbians.
Other key findings include: (a) respondents with conservative religious
affiliations had significantly higher levels of sexual prejudice than respondents with
liberal religious affiliations, no religious affiliation or non-religious respondents; (b)
Born-again Christian respondents had the highest level of sexual prejudice; (c)
compared to Born-again Christian respondents, Jewish respondents and respondents
identifying as spiritual with no specific religious affiliation had significantly more
favorable attitudes toward lesbians and gay men; (d) respondents who attend religious
services weekly or more often had significantly higher levels of sexual prejudice than
those who indicated they never attend, attend once to a few times each year and attend
one to three times each month; (e) women were close to more gay men than were men;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
and, (f) being close to one or more gay men in comparison with being close to no gay
men was associated with more favorable attitudes toward lesbians and gay men.
Twenty-one percent of the sample indicated images of gay men in the media
have in some way influenced their attitudes toward gay men, while 20% indicated
images of lesbians in the media have in some way influenced their attitudes toward
lesbians. When asked to rank the factors in terms of how they influence their attitudes
toward gay men and toward lesbians, 78% and 75% of the respondents respectively,
ranked personal values as first or second; 52% and 50% respectively, ranked
experiences with gay men and experiences with lesbians first or second; 47% and 51%
ranked religious beliefs first or second; 17% and 16% ranked views of friends and
family first or second; 17% ranked their definition of masculinity and 21% ranked their
definition of femininity as first or second; and 8% ranked media portrayals of gay men
and 9% ranked media portrayals as first or second.
Consistent with past findings, those with more conservative religious affiliations
and those who participate regularly in religious services tended to hold more
unfavorable attitudes toward gay men and lesbians. However, in this study, Catholic
respondents’ attitudes toward gay men and lesbians were more favorable than
Protestants’ attitudes, a result that was not expected. Although the European Catholic
faith consistently sends sexist and homophobic messages, American Catholicism may
slowly be evolving with American culture to be more accepting of gay men and
lesbians. It is also important to note that there could be varying degrees of liberalism
and conservativism within a denomination (e.g., some Catholics may view themselves
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
as being members of a liberal faith, while other Catholics may view themselves as
religiously conservative).
As hypothesized and in support of Allport’s (1954) contact hypothesis, being
close to gay men was associated with more favorable attitudes and being close to no gay
men was associated with less favorable attitudes. It is interesting that closeness to
lesbians was not associated with more favorable attitudes toward gay men and lesbians.
It is possible that some respondents in this sample had more negative experiences with
lesbians they were close to, which would negate the generally positive influence close
relationships tend to have on attitudes toward gay men and lesbians.
This is the first study to separate closeness to gay men or closeness to lesbians.
Past studies have assessed closeness to gay men or lesbians as one variable. It may be
important to separate these groups to determine if closeness to gay men and closeness to
lesbians have different influences on attitudes and attitude functions. In this study there
were respondent gender differences in closeness to gay men and lesbians. Women
tended to be close to more gay men than their male counterparts. These results may
suggest that since women are closer to more gay men, women’s attitudes are more
favorable than are men’s and the experiential-schematic function predicts more
favorable attitudes toward gay men and lesbians, in order for attitudes toward gay men
and lesbians to improve, gay men need to be open about their sexual orientation to
heterosexual male family members and co-workers.
Herek and Capitanio (1996) found that being close to two or more gay men or
lesbians was associated with more favorable attitudes, but not being close to one gay
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
man or lesbian. This study found that being close to one or more gay men significantly
improved attitudes. One explanation for this difference is that more of the respondents
in Herek’s study who were close to one gay man or one lesbian had more variable
experiences, meaning some negative and some positive, with the gay man or lesbian
they were close to. These varying attitudes toward gay men and lesbians may have led
to no significant differences between being close to no gay men or lesbians and being
close to one gay man or one lesbian.
Hypothesis lb: Heterosexual men will have less favorable attitudes toward gay
men and toward lesbians than heterosexual women. Most studies assessing attitudes
toward homosexuality have not separated gay men and lesbians into two distinct target
groups. The results of these studies have been inconsistent, but there has been a
growing body of research suggesting heterosexual women hold more favorable attitudes
toward homosexuality.
Consistent with Herek and Capitanio’s (1996) and Herek’s (2000) findings and
in support of this hypothesis, there were significant differences between heterosexual
men’s and heterosexual women’s attitudes toward gay men and toward lesbians.
Heterosexual women’s attitudes toward lesbians and attitudes toward gay men were
significantly more favorable than heterosexual men’s attitudes toward lesbians and
attitudes toward gay men. Heterosexual men’s attitudes toward gay men were the least
favorable. Heterosexual women’s attitudes toward lesbians were the most favorable,
while heterosexual men’s attitudes toward gay men were the least favorable.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
Herek (2002) hypothesized that differences in heterosexuals’ attitudes toward
lesbians and gay men should result primarily from differences in how heterosexual men
view lesbians compared to gay men. He predicted there would be a stronger negative
evaluation of homosexuals of one’s own sex for heterosexual men, yet not for
heterosexual women. He suggested that when the function of an attitude is to
demonstrate one’s masculinity or femininity, and in turn, their heterosexuality, sexual
prejudice will most likely be directed at the subgroup of homosexuals most closely tied
to their own sexual and gender identity, homosexuals of one’s own sex. The stronger
negative evaluation of homosexuals of one’s own gender is motivated by the desire not
to be confused as belonging to this particular group. There is a much greater concern
for men than for women in American society to demonstrate one’s heterosexuality, and,
at the same time, one’s gender role conformity. The results of this study supported
Herek’s (2002) hypothesis. Heterosexual women’s attitudes toward lesbians and
toward gay men were almost equivalent. Whereas, heterosexual men’s attitudes toward
gay men were significantly less favorable than their attitudes toward lesbians.
A possible explanation for why heterosexual women’s attitudes toward lesbians
are relatively favorable may have to do with two factors: elimination of competition and
potential stereotypes associated with lesbians. Heterosexual women may appreciate
lesbians for decreasing the competition for eligible heterosexual male mates, ultimately
leading to more acceptance of lesbians. Further, when heterosexual women think about
lesbians, they may retrieve a schema that represents women as strong, assertive and
independent, characteristics generally associated with men and ones that are highly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
valued in American culture. Therefore, the negative stereotypes of lesbians (e.g.,
manly, unattractive) may be outweighed by the positive ones leading to more favorable
attitudes toward lesbians. These potential explanations require empirical validation.
Hypothesis 1c: There will he a significant relationship between the number of
gay men and lesbians someone is close to and the defensive function. This hypothesis
was supported to the extent of closeness to gay men. There were significant differences
for the defensive function for gay men between respondents who were close to no gay
men and those who were close to two or more gay men. These findings suggest that
being close to no gay men increases endorsement of the defensive function. Closeness
to lesbians was not significantly associated with the defensive function. Fifty-eight
percent of respondents reported being close to no gay men.
One explanation for these findings is that respondents who are close to no gay
men, in this study primarily heterosexual men, may have unresolved sexual and gender
identity issues. These issues may lead them to experience extreme emotions, such as
revulsion when they think about lesbians and gay men. These extreme feelings and
thoughts may be one of the primary psychological motivations for why they hold their
attitudes toward gay men and lesbians. Because male homosexuality is more
prohibitive in American society, possibly because gay men are often equated with
feminine qualities that are not highly valued, the tendency to experience discomfort and
revulsion, two psychological factors associated with the defensive function, is
increased. In contrast, lesbians are many times linked with highly valued masculine
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
characteristics, therefore closeness to lesbians may be less related to the defensive
function.
Hypothesis Id; There will be gender group differences for the defensive
function. This hypothesis was supported. Heterosexual men and women differed
significantly on the defensive function as well as on the experiential-schematic function.
Heterosexual women’s experiential-schematic function scores for gay men were
significantly higher than heterosexual men’s scores on this same sub-scale. These
results suggest that heterosexual women tend to be psychologically motivated to hold
their attitudes toward gay men based on their experiences with gay people more than
heterosexual men do. In contrast, heterosexual men scored significantly higher than
heterosexual women on the defensive function for men. This suggests that heterosexual
men rely more on their revulsion of and discomfort with gay men and lesbians to guide
their attitudes toward gay men. Again, this supports Herek’s theory that heterosexual
men have more of a societal mandate to prove their heterosexuality and to deny any
feminine characteristics they may have, leading them to reject gay men through feelings
of revulsion and disgust. Interestingly, this need to be masculine because it is so highly
valued in American culture is also the same mechanism or dynamic that leads to
sexism. Feminine characteristics are devalued (e.g., cooperation versus competition,
soft versus hard, etc.) in the culture, which can lead to the continual promulgation of
sexism, sexual prejudice and patriarchy.
These results are important for many reasons. Since the experiential-schematic
function was found to be associated with more favorable attitudes toward gay men,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
there is a clear need for gay men and lesbians to come out to heterosexual men and
women. Gay men coming out to heterosexual men would presumably have the highest
favorable impact on attitudes toward gay men. This process could slowly dismantle
heterosexual men’s reliance on defensive psychological motivations when assessing
their feelings about gay men in the future.
However, these same men who have the defensive function present may also be
outwardly hostile to gay men and lesbians, making it very difficult for them to feel safe
enough to open up to these men. Gay men and lesbians may not take the risk to come
out to family members for fear of disownment. Co-workers may not come out because
of a lack of legal protection in the workplace. This is an important reason why the
Employment Non-Discrimination Act (ENDA) is such a critical bill to have passed.
Other workplace policies supportive of the basic rights for gay men and lesbians in the
workplace are also important.
Hypothesis 2: Heterosexual men will oppose civil rights for gay men and
lesbians more frequently than will heterosexual women. As hypothesized, there was
strong evidence to suggest that heterosexual men are less likely than women to agree
that lesbians and gay men should have legal rights and protection for marriage,
adoption, job prejudice and discrimination and serving openly in the military.
These results also suggest that there have been significant increases in support
for lesbian and gay rights since 1977, when some of the first polls on gay and lesbian
issues were administered. There have even been significant advances since 2000. For
example, in 2000, 39% of those surveyed agreed that same-sex couples should be able
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
to legally wed. la 2003, the percentage increased by 9% to 48%. Only five years ago
in a study asking the same question (cited in The Henry J. Kaiser Family Foundation,
2001), 33% of the respondents agreed with this statement.
In 1994, 29% of those polled (cited in The Henry J. Kaiser Family Foundation,
2001) agreed that gay men and lesbians should be allowed to adopt children. In 2000,
46% of those surveyed agreed that gay men and lesbians should be allowed to adopt
children. Three years later this increased by 11% to 57%.
In 1977, 56% of those polled believed that lesbians and gay men should be
legally protected from prejudice and discrimination in employment (cited in Rogers,
1998). It took over 20 years to realize an increase of 20% when the same question was
asked in 2000 (The Henry J. Kaiser Family Foundation, 2001). It took three more years
to increase agreement with this statement by 4% to 80%. Only 36% of the respondents
in this study disagreed with this statement.
The smallest gain was recorded for those who agree that lesbians and gay men
should be legally protected from prejudice and discrimination in employment, an
increase of 4% from 2000 (76%) to 2003 (80%). Interestingly, the highest percentage
gain, 13%, over the three-year period was for lesbians and gay men to be able to serve
openly in the military. Agreement went from 56% in 2000 to 69% in 2003.
When analyzing these results in the context of the geopolitical climate now and
just three years ago, there are some intriguing possible explanations. If these
explanations are ultimately empirically validated, they would add more evidence to
support Tourangeau and Rasinski’s (1988) claim that attitudes are many times fluid,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
changing evaluations that are dependent upon contextual factors such as political
climate, order of stimuli and even the mood of the individual when asked to assess their
thoughts and feelings about gay men and lesbians.
One possible explanation as to why there is more support for gay men and
lesbians to serve openly in the military is based on supply and demand. In 2000, when
56% of the population supported open service for gay men and lesbians, the United
States was in a relative state of peace. When the 2003 survey data was collected, the
influence of the terrorist attacks on September 11, 2001 still reverberated throughout the
United States. The country had just completed war in Afghanistan and was just
beginning to untangle itself from war in Iraq. The need for service members protecting
American soil was clearly evident. In this context, people’s negative feelings about
lesbians and gay men serving in the military may have been neutralized by a call to
band together. So, gay men and lesbians may no longer have been seen as the “other”
like Iraqis, Osama bin Laden, Muslims and Saddam Hussein. Instead, gay men and
lesbians serving in the military may have been perceived as Americans with a common
cause, a common enemy.
Agreement that lesbians and gay men should be legally protected from prejudice
and discrimination in employment increased 3% over that same period to 80% in 2003.
One possible explanation for the small percentage gain over the past three years is that
the economy has shown consistent recessionary signs since 2000, when 77% of those
surveyed supported job protection for gay men and lesbians. When asked in 2003,
many heterosexual Americans may have felt insecurity in their own employment
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
prospects. Therefore, legal job protections for gay men and lesbians became a low
priority or even something that caused anxiety when they considered this statement in
the context of the job market.
It is also interesting to note that fully 11% more heterosexual Americans
endorsed protection from job discrimination than heterosexual Americans who agreed
that lesbians and gay men should be able to serve openly in the military. This suggests
there may be biases in what jobs (e.g., military service, elementary school teacher, etc.)
heterosexual Americans believe gay men and lesbians should be protected from job
discrimination.
Agreement that gay men and lesbians should be able to adopt children increased
11% over the three-year period, from 46% to 57%. This increase could be due to the
visibility of gay and lesbian parents in the media over the past year. For example, in
February 2002, 14.4 million households watched “Rosie’s Story: For the Sake of the
Children” on Primetime Thursday. The two-hour special focused on Rosie O’Donnell’s
struggle to gain legal rights of gay and lesbian adoptive children. Also in 2002, “Nick
News Special Edition: My Family is Different” aired. The special looked at the
challenges faced by children of gay and lesbian parents. Some 1.6 million households
watched the episode, the highest audience numbers Nick News had over the past three
years (GLAAD, 2003a). These shows combined with Rosie O ’Donnell’s celebrity
stature may very well have increased the amount of public discourse on the topic of gay
and lesbian adoptions and gay and lesbian families. Further, Rosie O ’Donnell’s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
extraordinary popularity may have challenged some to revise their attitudes about gay
men and lesbians in general and their attitudes about gay and lesbian adoption.
Almost 10% fewer respondents agreed that same-sex couples should be able to
legally wed. This presents an interesting scenario. It may be acceptable for gay men
and lesbians to adopt children. But, for some of the same people who support gay and
lesbian adoption, they also believe same-sex relationships should not be legally
codified. This suggests that the notion of what family means in the United States may
have not changed very much since the 1950s when “family” meant a father, a mother
and children, leaving out anyone else who was primarily responsible for raising children
in a household.
Another possible explanation for the difference in opinions about adoption and
same-sex marriages may have to do with the wording of the question. More
respondents may believe that same-sex couples should have the same legal rights as do
heterosexual couples. However, some of these same individuals may feel strongly that
“marriage” is a religious institution that is reserved for heterosexual couples. The use
of the word “wed” in the question may have elicited a schema of marriage taking place
in a religious institution. The results may have been different if the question was
worded, “Same-sex couples should have the same legal rights and protections as do
married couples.” However, due to the use of the word “married” in this alternative
statement, the results may have been similar due to the affective and cognitive
responses “married” may elicit in heterosexuals.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
On the other hand, from 2000 to 2003 there was a 9% increase in the number of
respondents who did agree that same-sex couples should be able to legally wed, and for
possibly the first time since attitudes toward gay men and lesbians have been measured,
a majority, 48% to 45%, supported same-sex marriage rights.
Hypothesis 3a: The defensive function will predict less favorable attitudes
toward lesbians and toward gay men. This study presents evidence to suggest that
attitude functions can predict attitudes toward gay men and attitudes toward lesbians.
As hypothesized, the defensive function for both heterosexual men and women
predicted less favorable attitudes toward lesbians and toward gay men.
Interestingly, the defensive function for gay men—respondents were asked if
they based their attitudes toward gay men on two items—was the best predictor for
attitudes toward gay men as expected. However, the defensive function for gay men
was also the best predictor of attitudes toward lesbians for both heterosexual women
and men. Over 30% of the variance for heterosexual women’s attitudes toward lesbians
was accounted for by the defensive function for gay men. One explanation for this
occurrence is that even though heterosexual women and men were specifically asked
why they hold their attitudes toward lesbians, they were still thinking about gay men
from previous attitude scales where they were asked about their attitudes toward gay
men and toward lesbians. Therefore, the schema of lesbian would be subsumed into the
schema of gay man so that men and women would be actually thinking about gay men
when answering questions about what motivates their attitudes toward lesbians.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
The defensive function for lesbians was also a predictor of heterosexual men’s
attitudes toward lesbians, accounting for 2.1% of the variance explained. The defensive
function for lesbians also accounted for 2.9% of the variance explained for heterosexual
women’s attitudes toward lesbians and 2.9% of their attitudes toward gay men. One
explanation for this is heterosexual women may think about their attitudes toward gay
men even when asked about their attitudes toward lesbians because heterosexual
women are closer to more gay men than they are to lesbians, creating a response bias.
Hypothesis 3b: The social-expressive function will predict less favorable
attitudes toward lesbians and toward gay men. This hypothesis was not supported.
Only 1.2% of the variance for heterosexual women’s attitudes toward gay men was
explained by the social-expressive function for lesbians. The direction of the Beta
suggested that heterosexual women’s attitudes toward gay men were more favorable
when they were associated with the social-expressive function for lesbians.
These findings are important for two reasons. First, the theory expressed by
Haddock, Zanna, and Esses in 1993 suggesting that the social-expressive function will
predict unfavorable attitudes toward gay men and lesbians may actually have been true
10 years ago. However, with the improvement of attitudes toward gay men and lesbians
since then, high self-monitors or those who rely on others’ opinions to inform their own,
may tend to express more favorable attitudes toward gay men and toward lesbians. This
could explain why in this study the results indicated that at least for heterosexual
women, the social-expressive function predicted more favorable attitudes toward gay
men. Heterosexual women may perceive gay men to have characteristics (e.g., sense of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
style and appreciation for fashion similar to their own), making them in a sense a part of
their in-group. Heterosexual women's social groups may appreciate certain aspects of
gay men, explaining why this attitude function is predictive of attitudes toward gay
men.
Second, these findings are important because one of heterosexual women’s
motivations for holding attitudes toward gay men was the social-expressive function for
lesbians. This suggests that when women indicated they were motivated to hold their
attitudes toward lesbians by the attitudes of the people they are close to, this function
was ultimately associated with their motivation behind their attitudes toward gay men,
but not toward lesbians.
Hypothesis 3c: The value-expressive function will predict less favorable
attitudes toward lesbians and toward gay men. This hypothesis was not supported. The
value-expressive function scores were the highest of all four of the attitude functions,
meaning it was the predominant function responsible for the respondents’ attitudes.
However, the value-expressive function did not predict attitudes toward gay men or
attitudes toward lesbians. One explanation for the lack of predictive power for the
value-expressive function may be found in the two items that measure this sub-scale.
One of the items is based on protection of civil liberties for all people. The other item
taps respondents’ moral beliefs as the motivation for their attitudes toward gay men and
lesbians. Favorable and unfavorable attitudes toward gay men and lesbians could easily
be associated with these two items, thus explaining the lack of significance in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
relationship between the value-expressive function and attitudes toward gay men and
toward lesbians.
Because respondents’ highest average attitude function sub-scale score was for
the value-expressive function, it is important that future studies examine the API
(Herek, 1987) to determine if this sub-scale is possibly measuring two or more attitude
functions.
Hypothesis 3d: The experiential-schematic function will predict more favorable
attitudes toward lesbians and toward gay men. This hypothesis was supported. Similar
to the defensive function, the experiential-schematic function predicted heterosexual
men’s and heterosexual women’s attitudes toward lesbians and attitudes toward gay
men. In this study, the experiential-schematic function was the second strongest
predictor of attitudes and was associated with more favorable attitudes toward gay men
and toward lesbians.
Interestingly, heterosexual men’s attitudes toward gay men were predicted by
the experiential-schematic function for lesbians and heterosexual women’s attitudes
toward lesbians were predicted by the experiential-schematic function for gay men.
One explanation for these findings is that heterosexual men who had more favorable
attitudes toward gay men read the experiential-schematic function for lesbians as, “My
opinions about gay men are based mainly on whether or not someone I care about is
lesbian.” and “My opinions about gay men are based mainly on my personal
experiences with specific lesbians.” instead of the way in which the items were actually
worded, which was “My opinions about lesbians are based mainly on whether or not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
someone I care about is gay.” and “My opinions about lesbians are based mainly on my
personal experiences with specific gay persons.” This response bias could have
occurred because of male and female respondents’ increased comfort with thinking
about homosexuals of the other sex as opposed to ones of their own sex. This
interpretation should be empirically tested in future studies using a qualitative and
quantitative research design.
Hypothesis 4a: Heterosexual men’ s attitudes toward lesbians will be more
unfavorable when questions about gay men are asked first. This hypothesis was not
supported. Sex and order in which the questions were asked did not significantly
influence attitudes toward lesbians or attitudes toward gay men. However, item order
differences did approach significance. Overall, when questions about gay men were
asked first, heterosexual men’s attitudes toward gay men were the most unfavorable
followed by their attitudes toward lesbians. Further, assessing attitudes toward gay men
first led to the most unfavorable attitudes toward gay men and toward lesbians for all
respondents, regardless of respondent sex.
Since attitude survey research did not begin separating gay men and lesbians
into two target groups until the 1980s, it is difficult to know how HTV/AIDS has
influenced attitudes toward gay men. It would be important to empirically test how
much negative feelings toward gay men are associated with HIY/AIDS, feminine
qualities and revulsion by gay male sex. It would also be important to distinguish
differences between heterosexual men’s and heterosexual women’s specific reasons for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
their unfavorable attitudes toward gay men. This information could be useful in
devising attitude change strategies in the future.
When questions about lesbians were asked first, heterosexual women’s attitudes
toward lesbians were the most favorable, followed by heterosexual women’s attitudes
toward gay men. These results are interesting because the defensive function accounted
for more variance explained for heterosexual women’s attitudes toward both gay men
and toward lesbians than did the defensive function for heterosexual men’s attitudes.
This study found that attitudes motivated by the defensive function tend to be more
unfavorable and that heterosexual men tend to have more unfavorable attitudes toward
gay men and toward lesbians than do their female counterparts. It was expected that the
defensive function would have explained more of the variance for heterosexual men’s
attitudes toward gay men.
One explanation for this finding is that heterosexual women endorsed the
defensive function item about one’s desire to not think about homosexuality or gay
people more strongly than the other defensive function sub-scale item. This item could
be associated with more favorable attitudes toward lesbians and toward gay men. In
contrast, heterosexual men more strongly endorsed the other defensive function sub
scale item. This item inquired about one’s personal feelings of discomfort or revulsion
at homosexuality.
When questions about gay men were asked first, heterosexual men’s attitudes
toward lesbians were more unfavorable than when questions about lesbians were asked
first. The attitude scale mean difference was the highest in this case, suggesting that it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
is possible that heterosexual men's schema of lesbians and of gay men may be different.
This difference in the images that are retrieved when attitudes toward lesbians and
attitudes toward gay men are assessed, could account for the differences due to item
order. Because item order differences only approached significance, interpretation of
these results are being made cautiously.
As described by Herek (2002), when heterosexual men are asked about their
attitudes toward gay men and lesbians, they are assessing them through the sexual and
gender identity framework as opposed to a framework of minority politics. When asked
about their attitudes toward lesbians first, heterosexual men may think of lesbians in
terms of sexual pleasure and eroticism, thus explaining their more favorable attitudes
toward lesbians.
In contrast, when heterosexual men are asked about their attitudes toward gay
men first, they may experience feelings of disgust and revulsion toward homosexuality.
Some individuals may have had homosexual experiences in the past or have
homosexual desires. They may also feel like their heterosexuality is being threatened.
These same individuals may want to distance themselves from gay men because they
may also feel that these feelings and behaviors are wrong and are challenging to the
individual’s sense of self. They may also think of gay men as having undesirable
feminine characteristics. Until recently, many psychological associations equated
homosexuality with gender inversion. Gay men were presumed to be more like women
and lesbians were presumed to be more like men. Due to these potentially strong
unfavorable attitudes toward gay men, when questions about gay men are followed by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
questions about lesbians, heterosexual men may have carried over their schema of gay
men into their responses to questions about lesbians, thus explaining why their attitudes
toward lesbians in this condition were more unfavorable.
Hypothesis 4b: Heterosexual men will exhibit an increased defensive junction
when questions about gay men are asked before questions about lesbians. This
hypothesis was not supported because significant differences between sex and item
order by function were not found. However, on average the defensive function
increased for heterosexual men and less so for heterosexual women when questions
about gay men were asked before questions about lesbians.
Hypothesis 4c: Assessing attitudes about gay men first will make the defensive
function more accessible for male respondents, resulting in more unfavorable attitudes
toward gay men. Because no significant differences were found between item order and
sex by attitudes and by attitude functions, this hypothesis was not supported. However,
a comparison of average defensive function scores does show that asking questions
about gay men first led to the highest defensive function scores, with heterosexual men
scoring higher than heterosexual women.
The defensive function was associated with more unfavorable attitudes toward
gay men and toward lesbians. And, there were significant differences between
defensive function scores for attitudes toward gay men between heterosexual women
and heterosexual men. Heterosexual men scored higher than heterosexual women on
the defensive function sub-scale for attitudes toward gay men. This suggests that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
heterosexual men tend to use discomfort of and revulsion by gay men as primary
psychological motivators behind their unfavorable attitudes toward gay men.
Summary
This study was theoretically based in the functional (Katz, 1960) and neo
functional (Herek, 1987) approaches to attitudes; gender/sexual identity and minority
group paradigm models proposed by Herek (2000); context-dependent attitude theory
(Tourangeau & Rasinski, 1988); theory on media effects on worldview and value
system (Gross, 1991); and the contact hypothesis (Allport, 1954).
Overall, this research provided evidence that Herek’s (1987) neo-functional
model can be applied to better understanding the psychological motivation behind why
heterosexual men and women hold certain attitudes toward gay men and toward
lesbians. Evidence was also provided for his conceptual framework about how gender,
sexual identity and minority group paradigms help to explain differences in
heterosexual men’s and women’s attitudes toward lesbians and toward gay men.
Evidence was provided to support Allport’s (1954) contact hypothesis that states if one
can positively alter beliefs about a certain group through education or a positive
experience with an individual in the group, then discrimination will decrease and
attitudes toward that group will improve. Evidence supporting Gross’s (1991) theories
on the influence of the media on attitudes toward gay men and lesbians was not found.
However, determining media effects on attitudes toward gay men and lesbians was not
one of the primary variables examined in this study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
As noted In the procedure section, the data were self-reported and collected from
questionnaires completed through the Internet between June 26 and June 30 of 2003
with one reminder sent on June 29. It is of important note that on Thursday, June 26 the
Supreme Court of the United States handed down a decision in the Lawrence v. Texas
case. The Supreme Court found the Texas sodomy law to be unconstitutional, making
consensual sex between members of the same sex legal. This event was extensively
covered in the national and local media, making the results from this study especially
rich.
Studies have found that factors such as credibility and likeability of the
messenger significantly contribute to attitude change. The Supreme Court justices may
or may not score high on likeability scales. However, they would most likely receive
high marks for their credibility. The results from the four public policy questions in this
study suggest that the Supreme Court’s decision may have had a favorable influence on
the attitudes toward gay men and lesbians, at least temporarily. Interestingly, polls
conducted less than two months after the Lawrence v. Texas decision, but before
President Bush’s anti-gay marriage remarks and the Vatican’s stance on marriage being
only sanctioned between a man and a woman, showed declining support for the
legalization of gay marriage and other public policy issues affecting gay men and
lesbians.
The vast improvement of attitudes toward gay men and toward lesbians through
the decades may be due to an iterative process supported by Allport’s (1954) contact
hypothesis. As more gay men and lesbians reveal their gay and lesbian identities
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
publicly and more fair and inclusive representations of gay men and lesbians appear in
the media, family members, co-workers and friends of gay men and lesbians may be
confronted with cognitive dissonance (Festinger, 1957). Many resolve their conflicting
thoughts and feelings by revising and adopting more positive attitudes toward gay men
and lesbians as a group. These improved attitudes can manifest themselves through
greater support for gay men and lesbians in social, religious, educational and workplace
settings. This in turn, can lead to more gay men and lesbians coming out.
On the other hand, sometimes the mere hint of change in society can lead to
increased fear and threat, resistance to change and a tolerance for inequality by those
who tend to be politically conservative (Jost, Glaser, Kruglanski, & Sulloway, 2003).
There appears to have been a backlash toward rights for gay men and lesbians since the
Lawrence v. Texas decision. From the time the decision was announced in June to
September 2003, only three months later, hate crimes against gay men and lesbians
increased 52% (GLAAD, 2003b). This retreat in acceptance of gay men and lesbians
might be explained by public discourse about the possibility of gay and lesbian
legalized marriage, which began shortly after the decision. Americans might tend to
agree that what happens in the privacy of one’s home is something they could agree
with. But the idea of sharing the concept of legal marriage rights for gay men and
lesbians, which is so often related to religion in the minds of Americans, may threaten
their notion of the institution of marriage.
Whatever factors are responsible for influencing Americans’ attitudes toward
gay men and lesbians from May 7 through July 27,2003, there was a marked
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
fluctuation in attitudes. On May 5, a Gallup Poll (USA Today, 2003) indicated that 60%
of Americans polled believed that homosexual relations between consenting adults
should be legal. Less than 60 days later, a July 27 USA Toriay/CNN/Gallup Poll (USA
Today, 2003) found that only 50% of Americans polled agreed with this statement.
Fifty-seven percent of the respondents in this study agreed that gay men and lesbians
should be allowed to adopt children. Less than one month later, on July 27, in a USA
Torfay/CNN/Gallup Poll (USA Today, 2003), only 49% agreed that gay couples should
have the legal right to adopt a child.
These findings combined with the potential influence of history and the
geopolitical climate lends support to the context-dependent theory of Tourangeau and
Rasinski (1988). They suggested and have evidence supporting the theory that if
context plays a role in what attitude is expressed, each individual may have many
context-dependent attitudes and attitudes could possibly be constructed in an ad hoc
context-dependent manner. This implies that the commonly held belief that attitudes
are relatively long lasting and stable structures may not be accurate. Possibly, a
person’s attitude toward a social object may not be overwhelmingly favorable or
unfavorable, but may vary in degree based on what particular mental representation or
episode is activated and retrieved.
Findings from this study confirmed that gay men and lesbians should be
examined as separate target groups. Additionally, closeness to gay men and closeness
to lesbians should be assessed separately. Being female, close to one or more gay men,
of liberal or no religious affiliation and endorsing the experiential-expressive and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
social-expressive functions was associated with more favorable attitudes toward gay
men and toward lesbians. Being close to no gay men was associated with the defensive
function and was also associated with more unfavorable attitudes toward gay men and
lesbians. Heterosexual men tended to be close to no gay men more than their female
counterparts, leading to the conclusion that in order for heterosexual men’s attitudes
toward gay men and lesbians to improve, more openly gay men need to be close to
them.
Implications of the Study
Based on these results and theories, strategies to reduce sexual prejudice toward
gay men and lesbians could include: (a) more gay men and lesbians, especially gay
men, coming out to their heterosexual male family members and co-workers, (b)
psychologists and school counselors clearly understanding and being comfortable with
the dynamics between gender and sexual identity issues that many of their clients,
especially male clients, contend with, (c) increasing diverse gay and lesbian images in
films and television shows that target heterosexual males, such as Bravo’s “Queer Eye
for the Straight Guy,” which creates interactions between heterosexual men and gay
men, (d) those involved in religious organizations discussing the importance of
supporting the civil rights of gay men and lesbians with their religious leaders, (e)
parents teaching their children acceptance of and the value of diversity in American
culture, (f) gay men and lesbians becoming more involved in parents’ groups at their
children’s schools, (g) creating media programs that challenge the defensive function
and move viewers toward adoption of the experiential-schematic function, (h) creating
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
public education campaigns that convey messages that are respectful of and show how
supporting equality for Igbt people is congruent with their own personal values, (i)
parents and educators teaching boys new meanings of masculinity and girls new
meanings of femininity by shifting gender roles to more androgynous ones or by
expanding the boundaries of what it means to be a man and what it means to be a
woman, (j) creating opportunities for more accepting religious leaders to dialogue with
Born-again Christian and Baptist leaders about the importance of embracing their gay
and lesbian parishioners (k) the addition of gay and lesbian sensitivity trainings in all
employer diversity programs, especially for the military and (1) advocating for more
women in positions of influence throughout American culture, which could have a
significant positive influence on other’s attitudes.
In order to reduce heterosexual men’s tendency to be disgusted by gay men,
more reality television shows that encourage dialogue and interaction between gay men
and heterosexual men. This may result in increasing the knowledge heterosexual men
have about gay men and increase their utilization of the experiential-schematic function,
which is associated with more favorable attitudes toward gay men and lesbians.
Attitudes and attitude functions of both the contestants as well as audiences should be
assessed before the show, directly after the show and six to twelve months following the
show’s taping to determine if in fact there was an increase in the experiential-schematic
function and whether or not attitudes toward gay men and toward lesbians improved
and if that change lasted over time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
Further, since there are gender differences between closeness to gay men, with
men being close to fewer gay men, it is especially important for gay men to be open
about their sexual orientation to heterosexual men they interact with, which would most
likely lead to significantly improved attitudes toward gay men.
Further, to decrease endorsement of the defensive function, which is related to
more unfavorable attitudes toward gay men and lesbians, increasing images of physical
affection and sexual contact between people of the same sex in the media is
recommended. This could lead to normalizing same-sex physical contact in order to
shift defensive psychological states of denial and discomfort of homosexuality to more
neutral or positive attitudes.
Important messages to test in the media and with religious leaders would be
themes that illustrate that in general, gay men and lesbians share many of the same
values (family, education, honesty, loyalty, charity, community involvement and hard
work) that heterosexual Americans hold. With gay marriage front and center in the
political arena, it will be critical for images of loving and healthy gay and lesbian
families to be depicted in media images and stories.
Limitations o f the Study
There are many factors that contribute to the strength of this study’s
generalizability, such as the sample was relatively representative of the United States’
population. Most studies examining attitudes toward gay men, attitudes toward lesbians
or attitudes toward homosexuals as one target group have relied on paper and pencil
surveys with geographically anchored college student populations. This is the first
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
study of its kind to use an online survey method with a probability sample. This is also
the first study examining the influence of asking questions in a different order on
attitude functions.
When considering the interpretation of results and generalizability of this
study’s findings, limiting factors must be taken into consideration. Although this study
is unique because it used an adult sample whose demographic characteristics were
comparable to the adult population of the United States and the questionnaire was
administered through the Internet shielding the results from experimenter bias, there are
some internal reliability, external validity and measurement issues to consider.
One limitation stems from the self-report questionnaires that were used. Even
though respondents were informed their responses would be confidential and
anonymous when answering the questions while sitting at a computer, some
respondents might have veiled their true feelings and thoughts about homosexuality due
to the sensitive nature of the topic. Additionally, over 2,500 Harris Interactive panelists
were invited to participate in this study. Of those who self-selected into the study, 74
began completing the questionnaire and then dropped out. These potential respondents
may have dropped out of the study because they were uncomfortable with the survey
topic of attitudes toward lesbians and gay men. Or, they may have stopped for other
reasons such as time constraints. Without their completion of the survey and integration
of their data into the study, the results may or may not have been influenced.
Although this study used a random sample of over three million adults who
voluntarily enrolled in a web-based program that conducts numerous surveys on a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148
variety of topics, it is still not entirely representative of the United States population.
However, to select a sample that had similar characteristics of the United States
population, the variables of sex, geographic location and race were census data
controlled. One group was over-represented in this study’s sample—White
respondents. Due to the small number of respondents from non-White racial groups,
race-based comparisons could not be made in this study.
Of the measures used in this study, the ATL and ATG have been used in many
studies and have yielded acceptable to strong reliabilities. The AFT has usually been
used to measure attitude functions for attitudes toward homosexuals as one target group.
This study modified the wording of API questions to inquire about respondents’
perceptions of why they hold their attitudes toward lesbians and why they hold their
attitudes toward gay men. The results of the value-expressive sub-scale need to be
interpreted with caution due to possible concerns about its internal consistency.
Although the internal consistency reliabilities of the AFT subscales in this study were
comparable to past findings, results from the AFT-L and AFI-G warrant further
examination to determine if the AFT is applicable to lesbians and gay men as separate
target groups.
Additionally, history effects may have influenced the results of this study. The
study was deployed on June 26, 2003 the same day the Supreme Court of the United
States ruled that laws prohibiting same-sex sodomy were unconstitutional. Some
respondents may have been angered by the court’s decision and responded more
negatively to questions about gay men and lesbians. Other respondents who were more
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
ambivalent about their feelings toward these target groups may have decided that since
the United States Supreme Court Justices were in favor of specific civil rights for gay
men and lesbians, they too would take this anti-discrimination stance in their answers to
questions about gay men and lesbians. On the other hand, this same limitation may also
be a strength of the study. These results combined with other recent gay and lesbian
attitude surveys and polls, challenges the widely held assumption that attitudes are
stable and enduring schemas.
Future Directions
Due to the increase in violence toward gay men and lesbians, better
understanding the architecture of attitudes toward gay men and lesbians as well as the
psychological motivations supporting these attitudes is crucial. The results of this
study lay the groundwork for future studies that examine what specific messages and
interventions will be effective in improving heterosexual men’s and women’s
unfavorable attitudes toward gay men and toward lesbians. Additionally, to further
extend the results of this study, examining strategies to shift those who hold the
defensive function to functions associated with more favorable attitudes toward gay
men and lesbians (e.g., experiential-schematic function) is warranted.
Additionally, to better understand how heterosexual women and men organize
their attitudes toward gay men and lesbians, further research utilizing a qualitative
approach is recommended. A qualitative design would allow important questions to be
answered such as: (a) Do heterosexual women tend to organize their attitudes toward
gay men and lesbians in terms of a minority group paradigm that does not differentiate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150
between gay men and lesbians? And, (b) When attitudes toward gay men and lesbians
differ by respondent sex, is the difference due to evaluations based on a minority group
paradigm, sexualization of a target group, religious or political values or gender and/or
sexual orientation factors? “By asking anew of what use to people are their opinions,
such research has the potential to enrich our present understanding of how attitudes
form, and why they endure and change” (Herek, 1987, p. 301).
An ideal research design would include case studies and in-person interviews to
answer the remaining questions about why some respondents hold favorable attitudes
toward gay men and toward lesbians and why some hold unfavorable ones. Further, a
longitudinal research design would allow researchers to more accurately examine the
stability of both attitudes toward gay men and attitudes toward lesbians as well as
attitude functions among the same group of subjects over time.
Understanding differences in attitudes toward gay men and lesbians by race and
religion is also important. This study’s sample was primarily comprised of White
respondents, leaving an opportunity for future studies to focus on attitudes and attitude
functions of Asian-Americans, African-Americans, Hispanics as well as other racial
groups. This is important because there appears to be a growing gap between attitudes
toward gay men and lesbians along race and religious lines (Toner, 2003). Future
studies need to further examine these differences so that attitude change strategies will
factor in cultural and religious nuances that may make one strategy more effective than
another.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151
It appears there is a strong relationship between contact with and closeness to
gay men and favorable attitudes toward gay men and lesbians. Future studies that
examine what factors in a gay man’s environment influence his decision to come out or
not could inform psychologists, teachers, school counselors and human resources
professionals to advocate for the creation of optimal environments for their gay and
lesbian clients, students and employees. It is also important to determine why being
close to one or more gay men tends to lead to more favorable attitudes toward gay men
and lesbians. Conversely, it is important to examine why there are not gender
differences between being close to lesbians and attitudes toward gay men and lesbians.
The results from this study indicate that the value-expressive function registered
the highest mean score for heterosexual men and heterosexual women out of the four
attitude functions. This suggests that men and women identified their values as the
primary motivation for holding their attitudes toward gay men and lesbians. However,
the value-expressive function was not found to be a predictor of attitudes toward gay
men and lesbians. One explanation for this lack of predictive significance is that the
items in the AFT measuring the value-expressive function may in fact be measuring two
or more separate functions instead of one function. This explanation may be supported
by the comparably lower than expected internal consistency of the value-expressive
function in this study. The value-expressive items, “My opinions about lesbians and
gay men are based mainly on my concern that we safeguard the civil liberties of all
people in our society.” and “My opinions about lesbians and gay men are based mainly
on my moral beliefs about how things should be.” may tap two distinct factors. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
first item may tap the value of egalitarianism, while the second item may better measure
values based on homosexual prohibitive religious beliefs.
Future research could examine the value-expressive function to determine if
there may be a value-expressive (religious accepting), value-expressive (religious
prohibitive), value-expressive (equality accepting), value-expressive (equality rejecting)
and other possible value related motivations. This same study could also examine the
other functions for directional factors. For example, the experiential-schematic function
could be separated into experiential-schematic (positive) and experiential-schematic
(negative) to delineate between the quality of experiences with gay men and lesbians.
A qualitative and quantitative design, much like Herek’s (1987) could be used to
accomplish this goal.
Evidence from this study suggests that the defensive function is the best
predictor of more unfavorable attitudes toward gay men and lesbians. Being close to no
gay men and being a heterosexual male are associated with an increased defensive
function and more unfavorable attitudes toward gay men. Future studies could test what
messages and strategies are most effective for those who prominently hold the defensive
function in changing attitudes toward gay men and lesbians. Due to inconsistent
findings of studies in the past, more research needs to be conducted testing the matching
versus non-matching approach to attitude change. This study could provide information
about whether or not an argument or message that matches the function (e.g., social-
expressive function message could be that other members of their church group hold
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
153
favorable attitudes toward gay men and lesbians) is effective in changing attitudes
toward gay men and lesbians.
Conclusion
Sexual prejudice is not natural. It is taught. People can be taught otherwise.
This study focused on the roles of sex, age, religion, closeness to gay men and lesbians,
media effects, residence, age and education in understanding adult heterosexuals’
attitudes toward lesbians, attitudes toward gay men, opinions about gay and lesbian
public policy issues and attitude functions. This study also examined whether or not
attitude functions predict attitudes toward lesbians and toward gay men. Finally, to
better understand the importance of context on attitudes toward gay men and lesbians
and attitude functions, approximately half of the respondents were asked to think about
their attitudes toward lesbians and psychological motivations for holding these attitudes
before they were asked identical questions about gay men. The other half were asked
the same questions about gay men first and then asked about lesbians to help determine
if perhaps the schema of one target group may actually be subsumed by another group
that was recalled first.
The key findings of this study include:
1. Heterosexual women tended to hold similar attitudes toward gay men and
toward lesbians, whereas heterosexual men were more likely to hold different
attitudes according to the sex of the target. Heterosexual women’s attitudes
toward gay men and lesbians were more favorable than heterosexual men’s
attitudes. Heterosexual women’s attitudes toward lesbians were the most
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
favorable and heterosexual men’s attitudes toward gay men were the least
favorable.
2. Heterosexual women were more likely than men were to agree that lesbians and
gay men should be able to legally wed, should be able to adopt children, should
be legally protected from job prejudice and discrimination and should be able to
serve openly in the military. Of these four public policy issues, serving openly
in the military had the highest percentage gained, 13%, since 2000 (The Henry J.
Kaiser Family Foundation, 2001).
3. Jews, Catholics, Protestants and those who indicated they were spiritual with no
religious affiliation held significantly more favorable attitudes toward gay men
and lesbians than did Born-again Christians who held the least favorable
attitudes.
4. Heterosexuals who were close to more than one gay man generally held more
favorable attitudes toward lesbians and gay men compared to heterosexuals who
were close to no gay men. Fifty-eight percent of respondents reported being
close to no gay men.
5. Heterosexuals close to no gay men endorsed an increased defensive function.
6. The defensive function was associated with less favorable attitudes toward gay
men and lesbians and the experiential-schematic function was associated with
more favorable attitudes toward gay men and lesbians.
7. Heterosexual women tended to endorse the experiential- schematic function
more than heterosexual men when asked about their psychological motivation
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
behind their attitudes toward gay men. Heterosexual men tended to endorse the
defensive function more than heterosexual women.
8. The defensive function was the best predictor of heterosexuals’ attitudes toward
lesbians and gay men. The experiential-schematic function was the second best
predictor of attitudes toward gay men and lesbians.
9. Heterosexual women’s and men’s sexual prejudice and attitude function scores
did not differ significantly by item order.
Findings from this study contribute toward building theoretical knowledge of the
formation and change of attitudes toward gay men and lesbians as separate target
groups by supporting the importance of including relevant variables such as attitude
functions, sex, religion, closeness to gay men and lesbians, personal values, age,
education, residence and the geopolitical climate when conceptualizing attitudes toward
gay men and lesbians. The neo-functional theory, gender and sexual identity and
minority group paradigms, context-dependent attitude theory and the contact hypothesis
were the theories used to provide the conceptual framework for this study.
The need to extend conceptual and empirical knowledge about attitudes toward
gay men and lesbians as separate groups should now be apparent. Only by examining
attitudes toward and attitude functions for gay men and for lesbians will a clearer
picture emerge that distinguishes similarities and differences. These insights can be
used to create effective strategies to reduce sexual prejudice.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
156
References
Abramowitz, S. I., & Murray, J. (1983). Race effects in psychotherapy. In J. Murray &
P. R. Abramson (Eds.), Bias in psychotherapy (pp. 215-255). New York:
Praeger.
Aguero, J. E., Bloch, L., & Byrne, D. (1984). The relationships among sexual beliefs,
attitudes, experience, and homophobia in J. De Cecco (Ed.), Homophobia: An
overview. New York: The Haworth Press.
Ajzen, I. (2001). Nature and operation of attitudes. Annual Review o f Psychology, 52,
27-58.
Allport, G. W. (1954). The nature of prejudice. Garden City, NY: Doubleday Anchor.
Anderson, D. S., & Kristiansen, C. M. (1990). Measuring attitude functions. The
Journal of Social Psychology, 130, (3), 419-421.
Andison, F. (1977). TV violence and viewer aggression: A cumulation of study results
1956-1976. Public Opinion Quarterly, 41, 314-331.
Asch, S. E. (1940). Studies in the principles of judgments and attitudes: II.
Determination of judgments by group and by ego standards. Journal of Social
Psychology, SPSSI Bulletin, 12, 433-465.
Asch, S. E. (1952). Social psychology. Englewood Cliffs, NJ: Prentice Hall.
Bargh, J. A., Chaiken, S., Raymond, P., & Hymes, C. (1996). The automatic evaluation
effect: Unconditional automatic attitude activation with a pronunciation task.
Journal o f Experimental Social Psychology, 32, 104-128.
Baskin, D., Bluestone, H., & Nelson, M. (1981). Ethnicity and psychiatric diagnosis.
Journal of Clinical Psychology, 37, 529-537.
Beck, A. (1976). Cognitive therapy and the emotional disorders. New York:
International Universities Press.
Bern, D. J. (1965). An experimental analysis of self-persuasion. Journal of
Experimental Social Psychology, 1, 199-218.
Bern, S. (1981). Gender schema theory: A cognitive account of sex-typing.
Psychological Review, 88, 354-364.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
15?
Blumenfeld, W. J.» (1992). Homophobia: How we all pay the price. Boston: Beacon.
Brehm, J. W. (1956). Post-decision changes in the desirability of alternatives. Journal
of Abnormal and Social Psychology, 52, 925-939.
Cacioppo, J. T., Petty, R. E., Kao, C. F., & Rodriguez, R. (1986). Central and
peripheral routes to persuasion: An individual difference perspective. Journal of
Personality and Social Psychology, 51, 1032-1043.
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of
source versus message cues in persuasion. Journal of Personality and Social
Psychology, 39 ,752-756.
Chaiken, S. (1987). The heuristic model of persuasion. In M. P. Zanna, J. M. Olson, &
C. P. Herman (Eds.), Social influence: The Ontario symposium, (Vol. 5, pp. 3-
39). Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Christensen, S., & Sorenson, L. M. (1994). Effects of a multi-factor education program
on the attitude of child and youth worker students toward gays and lesbians.
Child & Youth Care Forum, 23 (2), 119-133.
Clark, R. D., & Maass, A. (1988). Social categorization in minority influence: The case
of homosexuality. European Journal of Social Psychology, 18 (4), 347-364.
Collins, B. E., Kiesler, C. A., & Miller, N. (1969). Attitude change: A critical analysis
of theoretical approaches. New York: John Wiley & Sons.
Crawford, I., & Solliday, E. (1996). The attitudes of undergraduate college students
toward gay parenting. Journal of Homosexuality, 30 (4), 63-77.
Cuenot, R. G., & Fugita, S. S. (1982). Perceived homosexuality: Measuring
heterosexual attitudinal and nonverbal reaction. Personality and Social
Psychology Bulletin, 8, (1) 100-106.
D’Augelli, A. R. (1989a). Heterosexual’s attitudes toward lesbians and gay men:
Correlates and gender differences. Journal of Sex Research, 25,451-477.
D’Augelli, A. R. (1989b). Homophobia in a university community: Views of
prospective resident assistants. Journal o f College Student Development, 30,
546-552.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
158
Deaux, K., & Lewis, L. L. (1984). Structure of gender stereotypes: Interrelationships
among components and gender label. Journal o f Personality and Social
Psychology, 46, 991-1004.
DeBono, K. (1987). Investigating the social adjustive and value-expressive functions of
attitudes: Implications for persuasion processes. Journal o f Personality and
Social Psychology, 52, 279-287.
Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled
components. Journal o f Personality and Social Psychology, 56, 5-18.
Dovidio, J. F., Brigham, J. C., Johnson, B. T., & Gaertner, S. L., (1996). Stereotyping,
prejudice, and discrimination: Another look. In C. N. Macrae, C. Stangor, & M.
Hewstone (Eds.), Stereotypes and stereotyping (pp. 276-322). New York:
Guilford Press.
Eagly, A. H., & Chaiken, S., (1993). The psychology of attitudes. San Diego, CA:
Harcourt Brace.
Eagly, A. H., Mladinic, A., & Otto, S. (1994). Cognitive and affective bases of
attitudes toward social groups and social policies. Journal o f Experimental
Social Psychology, 30, 113-137.
Eiser, J. R. (1987). The expression of attitudes. New York: Springer-Verlag.
Emulf, K. E., Innala, S. M., & Whitam, F. I. (1987). The relationship between affective
and cognitive components of homophobic reaction. Archives o f Sexual
Behavior, 16 (6 ), 501-509.
Esses, V. M., Haddock, G., & Zanna, M. P. (1993). Values, stereotypes, and emotions
as determinants of intergroup attitudes. In D. M. Mackie & D. L. Hamilton
(Eds.), Affect, cognition, and stereotyping: Interactive processes in group
perception (pp. 137-166). San Diego, CA: Academic Press.
Fazio, R. H., Jackson, J. R., Dunton, B. C., & Williams C. I. (1995). Variability in
automatic activation as an unobtrusive measure of racial attitudes-A bonafide
pipeline? Journal of Personality and Social Psychology, 69, 1013-1027.
Fenigstein, Allan, Scheier, M. F., & Buss, A. H. (1975). Public and private self-
consciousness: Assessment and theory. Journal o f Consulting and Clinical
Psychology, 43, 522-521.
Festinger, L. (1957). A theory of cognitive dissonance. Row, Peterson: Evanston, IL.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
159
Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression formation from
category-based to individuation processes: Influences of information and
motivation on attention and interpretation. In M. P. Zanna (Ed.), Advances in
Experimental Social Psychology (Vol. 23, pp. 1-74). San Diego, CA: Academic
Press.
Gastardo-Conaco, M. C. (1990). Complex social categories. Dissertation Abstracts
International, 51, 3184-3185.
Gay & Lesbian Alliance Against Defamation (2003a). GLAAD 2002 annual report: A
look back at a year’s worth of action by the people o f GLAAD (pp. 24-27).
GLAAD Annual Report, June 2003.
Gay & Lesbian Alliance Against Defamation (2003b). Anti-gay violence media
resource kit. Retrieved October 20, 2003 from http://www.glaad.org.
Gilchrist, R. S. (1995). An investigation o f value accessibility and its role in the value-
attitude-behavior relationship. Unpublished doctoral dissertation, University of
Western Ontario, London, Ontario, Canada.
Glaser, J., Sulloway, F., Jost, J., & Kruglanski, A. (2003). Political conservatism as
motivated social cognition. Psychological Bulletin.
Gleser, Goldine, C., & Ihilevich, D. (1969). An objective instrument for measuring
defensive mechanisms. Journal o f Consulting and Clinical Psychology, 33,
51-60.
Graham, S., Weiner, B., Giuliano, T., & Williams, E. (1993). An attributional analysis
of reactions to Magic Johnson. Journal of Applied social Psychology, 23,
996-1010.
Gross, L. (1984). The cultivation of intolerance. In G. Melischek, K. Rosengren, & J.
Stappers, (Eds.), Cultural indicators: An international symposium (pp. 345-
364). Vienna: Austrian Academy of Sciences.
Gross, L. (1991). Out of the mainstream: Sexual minorities and the mass media.
Journal o f Homosexuality, 21 (1-2), 19-46.
Haddock, G., Zanna, M. P., & Esses, V. M. (1993). Assessing the structure of
prejudicial attitudes: The case of attitudes toward homosexuals. Journal of
Personality and Social Psychology, 65, (6 ) 1105-1118.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
160
Hastie, R., Schroeder, C., & Weber, R. (1990). Creating complex social conjunction
categories from simple categories. Bulletin of Psychonomic Society, 28,
242-247.
Henry J. Kaiser Family Foundation, The, (2001). Inside-OUT: A report on the
experiences o f lesbians, gays and bisexuals in America and the public’s views
on issues and policies related to sexual orientation. (1-12). Menlo Park, CA:
Author.
Herek, G. M. (1984a). Beyond “homophobia”: A social psychological perspective on
attitudes toward lesbians and gay men. Journal of Homosexuality, 1 0 ,1-21.
Herek, G. M. (1984b). Attitudes toward lesbians and gay men: A factor analytic study.
Journal of Homosexuality, 10, (1/2), 39-51.
Herek, G. M. (1986a). On heterosexual masculinity: Some psychical consequences of
the social construction of gender and sexuality. American Behavioral Scientist,
29 (5), 563-577.
Herek, G. M. (1986b). The social psychology of homophobia: Toward a practical
theory. Review of Law & Social Change, 14 (4), 923-934.
Herek, G. M. (1987). Can functions be measured? A new perspective on the functional
approach to attitudes. Social Psychology Quarterly, 50 (4), 285-303.
Herek, G. M. (1988). Heterosexuals’ attitudes toward lesbians and gay men: Correlates
and gender differences. The Journal o f Sex Research, 25 (4), 457-477.
Herek, G. M. (1993). Documenting prejudice against lesbians and gay men on campus:
The Yale sexual orientation survey. Journal of Homosexuality, 25 (4), 15-30.
Herek, G. M. (2000). Sexual prejudice and gender: Do heterosexuals’ attitudes toward
lesbians and gay men differ? Journal o f Social Issues, 56 (2), 251-266.
Herek, G. M. (2002). Gender gaps in public opinion about lesbians and gay men.
Public Opinion Quarterly, 66 (1), 1-25.
Herek, G. M., & Capitanio, J. P. (1996). “Some of my best friends”: Intergroup
contact, concealable stigma, and heterosexuals’ attitudes toward gay men and
lesbians. Personality and Social Psychology Bulletin, 22 (4), 412-424.
Herek, G. M., & Capitanio, J. P. (1999). Sex differences in how heterosexuals think
about lesbians and gay men: Evidence from survey context effects. Journal of
Sex Research, 36 (4), 348-360.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
161
Hewstone, M., Hassebrauck, M., Wirtfa, A., & Waenke, M. (2000). Pattern of
disconfirming information and processing instructions as determinants of
stereotype change. British Journal of Social Psychology, 39, 399-411,
Human Rights Campaign (2001). 2001 FBI Hate crimes statistics. Retrieved August
20, 2003 from
http://www.hrc.org/issues/hate%5Fcrimes/background/stats/stats2001.asp.
Human Rights Watch (2001). U.S. gets “failing grade” protecting gay students:
Widespread bullying, teacher indifference in U.S. schools. Retrieved August
20, 2003 from http://www.hrw.org/press/2001/05/lgbt0530.htm.
Johnson, B. T., & Eagly, A. H. (1989). The effects of involvement on persuasion: A
meta-analysis. Psychological Bulletin, 106, 290-314.
Jost, J. T., Glaser, J., Kruglanski, A., & Sulloway, F. J. (2003). Exceptions that prove
the rule—Using a theory of motivated social cognition to account for ideological
incongruities and political anomalies. Psychological Bulletin, 129 (3), 383-393.
Kashima, Y., Woolcock, J., & Kashima, E. S. (2000). Group impressions as dynamic
configurations: The tensor product model of group impression formation and
change. Psychological Review, 107, (4), 914-942.
Katz, D. (1960). A functional approach to the study of attitudes. Public Opinion
Quarterly, 24, 163-204.
Kawakami, K., Moll, J., Hermsen, S., Dovidio, J. F., & Russin, A. (2000). Just say no
(to stereotyping): Effects of training in the negation of stereotypic associations
on stereotype activation. Journal o f Personality and Social Psychology, 78, (5),
871-888.
Kilman, P. R., Wanlass, R. L., Sabalis, R. R., & Sullivan, B. (1981). Sex education: A
review of its effects. Archives of Sexual Behavior, 10, 177-205.
Kite, M. E. (1984). Sex differences in attitudes toward homosexuals: A meta-analytic
review. Homophobia: An overview. Special Edition. Journal of
Homosexuality, 69-81.
Kite, M. E., & Whitley, B. E., Jr. (1996). Sex differences in attitudes toward
homosexual persons, behaviors, and civil rights: A meta-analysis. Personality
and Social Psychology Bulletin, 22, 336-353.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
162
Krech, D., & Crutchfield, R. S. (1948). Attitudes and society. New York: McGraw-
Hill.
Kristiansen, C. M. (1990). The symbolic/value-expressive function of out-group
attitudes among homosexuals. The Journal of Social Psychology, 130, (1),
61-69.
Kristiansen, C. M., & Zanna, M. P. (1988). Justifying attitudes by appealing to values:
A functional perspective. British Journal of Social Psychology, 27, 247-256.
Krosnick, J. A., (1988). The role of attitude importance in social evaluation: A study of
policy preferences, presidential candidate evaluations, and voting behavior.
Journal o f Personality and Social Psychology, 55, 196-210.
Kuklinski, J. H., Riggle, E., Ottati, V., Schwarz, N., & Wyer, Jr., R. S. (1991). The
cognitive and affective bases of political tolerance judgments. American
Journal of Political Science. 36 (1), 1-27.
Lemyre, L., & Smith, P. M. (1985). Intergroup discrimination and self-esteem in the
minimal group paradigm. Journal of Personality and Social Psychology, 49,
660-670.
Linville, P. W., & Carlston, D. (1994). Social cognition of the self. In P. G. Devine, D.
L. Hamilton, & T. M. Ostrom (Eds.), Social cognition: Impact on social
psychology (pp. 143-193). San Diego, CA: Academic Press.
Littlewood, R. (1992). Psychiatric diagnosis and racial bias: Empirical and interpretive
approaches. Social Science and Medicine, 34, 141-149.
Lord, C. G., Lepper, M. R., & Mackie, D. (1984). Attitude prototypes as determinants
of attitude-behavior consistency. Journal o f Personality and Social Psychology,
4 6 ,1254-1266.
Luhrs, T., Crawford, I., & Goldberg, J. (1991). The presence of the defensive function
as a predictor of heterosexual college students’ affective responses toward gay
men and lesbians. Journal of Psychology & Human Sexuality, 4 (4), 71-87.
Mackie, D. M. (1998). Intergroup relations: Insights from a theoretically integrative
approach. Psychological Review, 105 (3), 499-529.
Macrae, C. N., Bodenhausen, G. V., & Milne, A. B. (1997). Saying no to unwanted
thoughts: The role of self-awareness in the regulation o f mental life.
Unpublished manuscript.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
163
Macrae, C. N., Bodenhausen, G. V., Milne, A. B., & Jetten, J. (1994). Out of mind but
back in sight: Stereotypes on the rebound. Journal of Personality and Social
Psychology, 67, 808-817.
Maio, G. R., & Olson, J. M. (1994). Value-attitude-behaviour relations: The
moderating role of attitude functions. British Journal o f Social Psychology, 33,
301-312.
Maio, G. R., & Olson, I. M. (1995). The effect of attitude dissimulation on attitude
accessibility. Social Cognition, 13, (2), 127-144.
Milavsky, J., Kessler, R., Stipp, H., & Rubens, W. (1982). Television and aggression:
The results of a panel study. New York: Academic Press.
Millar, M. G., & Millar, K. U. (1990). Attitude change as a function of attitude type
and argument type. Journal of Personality and Social Psychology, 59, (2),
217-228.
Millar, M. G., & Millar, K. U. (1996). The effects of direct and indirect experience on
affective and cognitive responses and the attitude-behavior relation. Journal of
Experimental Social Psychology, 32, 561-579.
Millar, M. G., & Tesser, A. (1986). Effects of affective and cognitive focus on the
attitude-behavior relation. Journal o f Personality and Social Psychology, 51,
(2), 270-276.
Miller, N., & Brewer, M. B. (1986). Categorization effects on in-group and out-group
perception. In J. F. Dovidio & S.L. Gaertner (Eds.), Prejudice, discrimination,
and racism (pp. 209-230). San Diego, CA: Academic Press.
Monteith, M., Sherman, I., & Devine, P. (1998). Suppression as a stereotype control
strategy. Personality and Social Psychology Review, 1, 63-82.
Murray, S. L., Haddock, G., & Zanna, M. P. (1996). On creating value-expressive
attitudes: An experimental approach. In C. Seligman, J. M. Olson, et al. (Eds.),
The psychology of values: The Ontario symposium (Vol. 8 , pp. 107-133).
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Nardi, P. M. (1997). Changing gay and lesbian images in the media. In J. T. Sears &
W. L. Williams (Eds.), Overcoming heterosexism and homophobia: Strategies
that work (pp. 49-64). New York: Columbia University Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
164
National Mental Health Association (2003). What does gay mean? How to talk with
kids about sexual orientation and prejudice. Retrieved August 21, 2003 from
http://www.nmha.org/whatdoesgaymean/.
Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A meta-analysis.
Psychological Bulletin, 144 (1), 29-51.
Ostrom, T. M., & Brock, T. C. (1969). Cognitive bonding to central values and
resistance to a communication advocating change in policy orientation. Journal
of Experimental Research in Personality, 4, 42-50.
Out & Equal (2002). Workplace summit breaks 12-year attendance record: Over 400
show up in Orlando for three-day conference to learn ways to advocate for
safety and equity. Out & Equal Workplace Advocates, 4 (4), 1.
Page, S. (2003, July 28). Gay rights tough to sharpen into political ‘wedge issued USA
TODAY, (p. 10A).
Petty, R. E., & Cacioppo, J. T. (1981). Attitudes and persuasion: Classic and
contemporary approaches. Dubuque, IA: William C. Brown.
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion.
In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 19, pp.
123-205). San Diego, CA: Academic Press.
Reardon, K. K. (1981). Persuasion: Theory and context. Beverly Hills, CA: Sage.
Ridley, C. R. (1986). Diagnosis as a function of race pairing and client self-disclosure.
Journal o f Cross-Cultural Psychology, 17, 337-351.
Rogers, M. A. (1998). Americans’ attitudes toward gays and lesbians since 1965.
Doctoral dissertation, American University, 1998, 1-391.
Rokeach, M. (1967). Value survey. Sunnyvale, CA: Halgren Tests.
Rudolph, J. (1990). Counselors’ attitudes toward homosexuality: A selective review of
the literature. Journal of Counseling and Development, 67 (3), 165-168.
Russo, V. (1987). The celluloid closet: Homosexuality and the movies. New York:
Harper and Row.
Ryder, N. B. (1965). The cohort as a concept in the study of social change. American
Sociological Review, 30 December, 843-861.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
165
Schaller, M. (1992). In-group favoritism and statistical reasoning in social inference:
Implications for formation and maintenance of group stereotypes. Journal of
Personality and Social Psychology, 63, 61-74.
Schiller, H. I. (1973). The mind managers. Boston: Beacon Press.
Sears, D. O. (1988). Symbolic racism. In P. A. Katz & D. A. Taylor (Eds.),
Eliminating racism: Profiles in controversy (pp. 53-84). New York: Plenum
Press.
Shavitt, S. (1989). Operationalizing functional theories of attitude. In A. R. Pratkanis,
S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp.
311-337). Hillsdale, NY: Erlbaum.
Shavitt, S. (1990). The role of attitude objects in attitude functions. Journal of
Experimental Social Psychology, 26, 124-148.
Sia, T. L., Lord, C. G., Blessum, K. A., Ratcliff, C. D., & Lepper, M. R. (1997). Is a
rose always a rose? The role of social category exemplar change in attitude
stability and attitude-behavior consistency. Journal of Personality and Social
Psychology, 72 (3), 501-514.
Smith, E. R. (1992). The role of exemplars in social judgment. In L. L. Martin & A.
Tesser (Eds.), The construction o f social judgments (pp. 37-65). Hillsdale, NJ:
Erlbaum.
Smith, E. R. (1994). Procedural knowledge and processing strategies in social
cognition. In r. S. Wyer, Jr., & T. K. Srull (Eds.), Handbook o f social cognition
(Vol. 1, pp. 99-151). Hillsdale, NJ: Erlbaum.
Smith, M. B., Bruner, J. S., & White, R. W. (1956). Opinions and personality. New
York: Wiley.
Snyder, M., (1974). The self-monitoring of expressive behavior. Journal of
Personality and Social Psychology, 30, 526-537.
Srull, T. K., & Wyer, R. S. (1989). Person memory and judgment. Psychological
Review, 96, 58-83.
Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the
self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol.
21, pp. 261-302). San Diego CA: Academic Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
166
Stevenson, M. R., & Gajarsky, W. M. (1990). Issues of gender in promoting tolerance
for homosexuality. Journal o f Psychology & Human Sexuality, 3 (2), 155-163.
Stewart, C. (1997). Sexual orientation training in law enforcement agencies: A
preliminary review of what works. In J. T. Sears & W. L. Williams (Eds.),
Overcoming heterosexism and homophobia: Strategies that work (pp. 326-338).
New York: Columbia University Press.
Stimson, J. A. (1991). Public opinion in America: Moods, cycles, and swings. Boulder:
Westview Press.
Strickland, T. L., Jenkings, J. O., Myers, H. F., & Adams, H. E. (1988). Diagnostic
judgments as a function of client and therapist race. Journal of
Psychopathology and Behavioral Assessment, 10, 141-151.
Tajfel, H. (1978). Differentiation between social groups: Studies in the social
psychology of inter group relations. New York: Academic Press.
Tesser, A. (1988). Toward a self-evaluation maintenance model of social behavior. In
L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp.
181-227). San Diego, CA: Academic Press.
Toner, R. (2003, July 25). Opposition to gay marriage is declining, study finds. The
New York Times, July 25, 2003, (p. A16).
Tourangeau, R., & Rasinski, K. A., (1988). Cognitive processes underlying context
effects in attitude measurement. Psychological Bulletin, 103, 299-314.
Triandis, H. C. (1971). Attitude and attitude change. New York: John Wiley & Sons.
U.S. Census Bureau, (2002). Census 2000. Retrieved July 23, 2003, from
http://www.census.gov.
Voss, J. R. (1980). Sex education: Evaluation and recommendations for future study.
Archives of Sexual Behavior, 9, 37-59.
Werth, J. L., & Lord, C. G. (1992). Previous conceptions of the typical group member
and the contact hypothesis. Basic and Applied Social Psychology, 13, 351-369.
Whitley, B. E., & Kite, M. E. (1995). Sex differences in attitudes toward
homosexuality: A comment on Oliver and Hyde (1993). Psychological Bulletin,
117(1), 146-154.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
167
Williams, W. L. (1997). Introduction. In J. T. Sears, & W. L. Williams (Eds.),
Overcoming heterosexism and homophobia (pp. 1-10). New York: Columbia
University Press.
Wilson, T. C. (1996). Cohort and prejudice, Whites’ attitudes toward Blacks,
Hispanics, Jews, and Asians. Public Opinion Quarterly, 60 (2), 253-274.
Wilson, T. D., Lisle, D. J., & Schooler, J. (1990). Some undesirable effects of self-
reflection. Unpublished manuscript, University of Virginia, Charlottesville.
Wood, W. (2000). Attitude change: Persuasion and social influence. Annual Review of
Psychology, 51, 539-570.
Wright, S. C., Aron, A., McLaughlin-Volpe, T., & Ropp, S. A. (1995). The extended
contact effect: Knowledge of cross-group friendships and prejudice. Journal of
Personality and Social Psychology, 73, 73-90.
Yeb, G. A. (1997). Changing homophobic and heterosexist attitudes: An overview of
persuasive communication approaches. In J. T. Sears & W. L. Williams (Eds.),
Overcoming heterosexism and homophobia: Strategies that work (pp. 49-64).
New York: Columbia University Press.
Yzerbyt, V., Coull, A., & Rocher, S. (1999). Fencing off the deviant: The role of
cognitive resources in the maintenance of stereotypes. Journal of Personality
and Social Psychology, 77,449-462.
Zuwerink, J. R., & Devine, P. G. (1996). Attitude importance and resistance to
persuasion: It’s not just the thought that counts. Journal o f Personality and
Social Psychology, 70 (5), 931-944.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
168
Appendix A
Today we’re conducting a survey on attitudes people have about different topics. First
we would like to ask you some questions for classification purposes, so that we may
customize the survey for you.
Are you...?
Please enter your age in the box below. Please round to the nearest year.
I J J years old
In what country or region do you currently reside?
rDRQP DOWN MENU WITH CHOICES LISTED -SEE STANDARD RESPONSES!
In what state or territory do you currently reside?
[DROP DOWN MENU WITH CHOICES LISTED -SEE STANDARD RESPONSES]
What is your zip code? Please enter only the first five digits.
In what province or territory do you currently reside?
53 Alberta
54 British Columbia
55 Manitoba
56 New Brunswick
57 Newfoundland
76 Northwest Territories
58 Nova Scotia
77 Nunavut
60 Ontario
61 Prince Edward Island
62 Quebec
1
2
3
Male
Female
Other gender
C ANADIAN RESPONDENTS
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
169
63 Saskatchewan
75 Yukon Territory
BASE: CANADIAN RESPONDENTS
What is your postal code?
BASE: NON-IJ.S./NON-CANADIAN RESPONDENTS
Q340 What is your postal code?
[IF NON-U.S. (Q315/NOT 244) TERMINATE TO Q399]
1 In an urban or city area
2 In a suburban area next to a city
3 In a small town or rural area
Do you consider yourself... ?
01 White
03 Hispanic
04 Asian or Pacific Islander
05 Native American or Alaskan native
06 Mixed racial background
07 Other race
08 African American
09 First Nation/Native Canadian
10 South Asian
12 Chinese
13 Korean
14 Japanese
15 Other Southeast Asian
16 Filipino
17 Arab/West Asian
94 Decline to answer
BASE: OTHER RACE
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
170
Q360 Please specify what race you consider yourself.
[TEXT BOX]
BASE; J J ^ E S P C M D E IT A N P MIXED RACIAL BACKGROUND
You indicated that you consider yourself of a mixed racial background. With which of
the following racial groups do you most closely identify? Please check all that apply.
01 White
02 Black
03 African American
04 Asian or Pacific Islander
05 Native American or Alaskan native
06 Other race
94 Decline to answer
Do you know anyone who is gay, lesbian, bisexual, or transgender? This can be
someone you are close to or someone who is an acquaintance. Please check all that
apply.
1 Yes, a family member
2 Yes, a close personal friend
3 Yes, a co-worker
4 Yes, a friend or acquaintance (not a co-worker)
5 Yes, another person not mentioned
6
7
No
Not sure
Do you consider yourself... ?
[RANDOMIZE 1-5.]
1
2
3
4
6
7
8
Heterosexual (straight)
Lesbian
Gay
Bisexual
Other
Not sure
Decline to answer
M SElQ TB ER ^X IIA ^
Please specify your sexual orientation.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
171
[TEXT BOX]
Q399 BEHIND THE SCENES - A QUALIFIED RESPONDENT MUST
MEET THE FOLLOWING CRITERIA:
-BE AGE 18+
-BE A U.S. RESPONDENT
-BE HETEROSEXUAL
We would now like to ask you a few background questions about your religious beliefs
(if any).
Please indicate your religious attendance in the past year.
1 Never
2 Once to a few times
3 1-3 times/month
4 Weekly or more often
9 Decline to answer
My religious affiliation is...?
1 Conservative (examples include Mormon, Catholic, Baptist, etc.)
2 Liberal (examples include Presbyterian, Episcopalian, Jewish reform,
etc.)
3 Other
4 No religious affiliation
5 Not religious
6 Decline to answer
Please indicate your religious affiliation.
Are you...?
1 Catholic
2 Jewish
3 Born-again Christian
4 Baptist
5 Buddhist
6 Protestant
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
172
7 Seventh Day Adventist
8 Agnostic
9 Other religion
10 Spiritual with no specific religious affiliation
11 Atheist
99 Decline to answer
How many gay men are you close to? By “close", we mean you speak to them on a
regular basis and are emotionally tied to them in some way.
1 0
2 1
3 2
4 3
5 4
6 5+
7 Decline to answer
How many gay men are you an acquaintance with? By “acquaintance ”, we mean
someone you know slightly, but are not emotionally tied to.
1 0
2 1
3 2
4 3
5 4
6 5+
7 Decline to answer
In general, have your experiences with gay men been...?
1 Favorable
2 Unfavorable
3 Both (had favorable experiences, as well as unfavorable experiences)
4 Neutral (neither favorable nor unfavorable)
5 Decline to answer
How many lesbians are you close to? By “close ”, we mean you speak to them on a
regular basis and are emotionally tied to them in some way.
1 0
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
173
2 1
3 2
4 3
5 4
6 5+
7 Decline to answer
How many lesbians are you an acquaintance with? By “ acquaintance ”, we mean
someone you know slightly, but are not emotionally tied to.
1 0
2 1
3 2
4 3
5 4
6 5+
7 Decline to answer
In general, have your experiences with lesbians been...?
1 Favorable
2 Unfavorable
3 Both (had favorable experiences, as well as unfavorable experiences)
4 Neutral (neither favorable nor unfavorable)
5 Decline to answer
Please indicate the number of hours, on average, you watch television on a daily basis.
I J J J hours a day
Please fill in the blank within the statement below:
Overall, the images of gay men that I have seen in the media over the past two years
have portrayed gay men in a __________light.
1 2 3 4 5
Positive Somewhat Positive Neither Positive Somewhat Negative Negative
Nor Negative
Please fill in the blank within the statement below:
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
174
Overall, the images of lesbians that I have seen in the media over the past two years
have portrayed lesbians in a __________ light.
1 2 3 4 5
Positive Somewhat Positive Neither Positive Somewhat Negative Negative
Nor Negative
Please complete the following statement:
Images of gay men in the media have influenced my attitudes toward gay men...?
1 Not at all
2 A little
3 Somewhat
4 A lot
5 Decline to answer
Please complete the following statement:
Images of lesbians in the media have influenced my attitudes toward lesbians... ?
1 Not at all
2 A little
3 Somewhat
4 A lot
5 Decline to answer
Please rank the following factors in terms of how they influence your attitudes toward
gay men. Please rank the factors with 1 being the most influential and 6 being the least
influential.
Please use 1, 6, and every number in between.
Rank
1 Media portrayals of gay men IJ
2 Religious beliefs 1J
3 Personal values IJ
4 Views of my friends and family l_l
5 Personal experiences with gay men
LI
6 My definition of masculinity IJ
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
175
Please rank the following factors in terms of how' they influence your attitudes toward
lesbians. Please rank the factors with 1 being the most influential and 6 being the least
influential. Please use 1, 6, and every number in between.
Rank
1 Media portrayals of lesbians LI
2 Religious beliefs IJ
3 Personal values IJ
4 Views of my friends and family l_l
5 Personal experiences with lesbians LI
6 My definition of femininity IJ
BEHIND THE SCENES QUESTION
Randomly place respondents into 1 of the following 2 groups. The group they are
placed in determines the order they will see Q515 and Q520, then Q525 and Q530.
Group 1 will see Q515 and Q520 (as a block) first, then theyTl see Q525 and Q530 (as
a block) second. Group 2 will see Q525 and Q530 (as a block) first, then Q515 and
Q520 (as a block) second.
Q515 Please use the numbers below to indicate how much you agree or disagree with
each statement. Please answer the questions as honestly as possible.
(Attitudes Toward Lesbians Scale)
Q516 1 2 3 4 5
Strongly Somewhat Neither Agree Somewhat Strongly
Disagree Disagree Nor Disagree Agree Agree
1 Lesbians just can’t fit into our society.
2 A woman’s homosexuality should not be a cause for job discrimination.
3 Female homosexuality is detrimental to society because it breaks down
the natural divisions between the sexes.
4 State laws regulating private, consenting lesbian behavior should be
loosened.
5 Female homosexuality is a sin.
6 The growing number of lesbians indicates a decline in American morals.
7 Female homosexuality in itself is no problem, but what society makes of
it can be a problem.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
176
8 Female homosexuality is a threat to many of our basic social institutions.
9 Female homosexuality is an inferior form of sexuality.
10 Lesbians are sick.
(Attitude Function Inventory-Lesbians)
Q520 Please use the numbers below to indicate how much you agree or disagree with
each statement. Please answer the questions as honestly as possible.
Q521 1 2 3 4 5
Strongly Somewhat Neither Agree Somewhat Strongly
Disagree Disagree Nor Disagree Agree Agree
1 My opinions about lesbians are based mainly on whether or not someone I care
about is gay.
2 My opinions about lesbians are based mainly on my personal experiences with
specific gay persons.
3 My opinions about lesbians are based mainly on my judgment of how likely it is
that I will interact with gay people in any significant way.
4 My opinions about lesbians are based mainly on my personal experiences with
people whose family members or friends are gay.
5 My opinions about lesbians are based mainly on my perceptions of how the
people I care about have responded to gay people as a group.
6 My opinions about lesbians are based mainly on learning how gay people are
viewed by the people whose opinions I most respect.
7 My opinions about lesbians are based mainly on the fact that I would rather not
think about the homosexuality of gay people.
8 My opinions about lesbians are based mainly on my personal feelings of
discomfort or revulsion at homosexuality.
9 My opinions about lesbians are based mainly on my concern that we safeguard
the civil liberties of all people in our society.
10 My opinions about lesbians are based mainly on my moral beliefs about how
things should be.
(Attitudes Toward Gay Men Scale)
Q525 Please use the numbers below to indicate how much you agree or disagree with
each statement. Please answer the questions as honestly as possible.
Q526 1 2 3 4 5
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
177
Strongly Somewhat Neither Agree Somewhat Strongly
Disagree Disagree Nor Disagree Agree Agree
1 Gay men just can’t fit into our society.
2 A man’s homosexuality should not be a cause for job discrimination.
3 Male homosexuality is detrimental to society because it breaks down the
natural divisions between the sexes.
4 State laws regulating private, consenting gay male behavior should be
loosened.
5 Male homosexuality is a sin.
6 The growing number of gay men indicates a decline in American morals.
7 Male homosexuality in itself is no problem, but what society makes of it
can be a problem.
8 Male homosexuality is a threat to many of our basic social institutions.
9 Male homosexuality is an inferior form of sexuality.
10 Gay men are sick.
(Attitude Function Inventory-Gay Men)
Q530 Please use the numbers below to indicate how much you agree or disagree with
each statement. Please answer the questions as honestly as possible.
Q531 1 2 3 4 5
Strongly Somewhat Neither Agree Somewhat Strongly
Disagree Disagree Nor Disagree Agree Agree
1 My opinions about gay men are based mainly on whether or not someone
I care about is gay.
2 My opinions about gay men are based mainly on my personal
experiences with specific gay persons.
3 My opinions about gay men are based mainly on my judgment of how
likely it is that I will interact with gay people in any significant way.
4 My opinions about gay men are based mainly on my personal
experiences with people whose family members or friends are gay.
5 My opinions about gay men are based mainly on my perceptions of how
the people I care about have responded to gay people as a group.
6 My opinions about gay men are based mainly on learning how gay
people are viewed by the people whose opinions I most respect.
7 My opinions about gay men are based mainly on the fact that I would
rather not think about the homosexuality of gay people.
8 My opinions about gay men are based mainly on my personal feelings of
discomfort or revulsion at homosexuality.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
178
9 My opinions about gay men are based mainly on my concern that we
safeguard the civil liberties of all people in our society.
10 My opinions about gay men are based mainly on my moral beliefs about
how things should be.
Please indicate your agreement or disagreement with the following statements.
Same-sex couples should be able to legally wed.
1 I agree with this statement.
2 I disagree with this statement.
3 Decline to answer
Gay men and lesbians should be allowed to adopt children.
1 I agree with this statement.
2 I disagree with this statement.
3 Decline to answer
Lesbians and gay men should be legally protected from prejudice and discrimination in
employment.
1 I agree with this statement.
2 I disagree with this statement.
3 Decline to answer
Lesbians and gay men should be able to serve openly in the military.
1 I agree with this statement.
2 I disagree with this statement.
3 Decline to answer
The next series of questions is primarily for classification purposes and will help us
properly analyze responses to this survey. As you may already know, we never disclose
the identity of any individual. Your answers will always be kept strictly confidential.
We only report results for groups of people, not for individuals.
What type of Internet connection do vou have for your home computer or other
prim ary computer?
01 14.4k modem
02 28.8k modem
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
179
03 33.6k modem
04 56k modem
06 Cable modem
07 T1 or T3 line
08 ISDN line
09 ADSL/DSL
96 Other
98 Not sure
What is your marital status?
1 Single, never married
2 Married
3 Divorced
4 Separated
5 Widowed
7 Partnered
How many children under the age of 18 live in your household?
What is the highest level of education you have completed or the highest degree you
have received?
1 Less than high school
2 Completed some high school
3 High school graduate or equivalent (e.g., GED)
4 Completed some college, but no degree
5 Associate’s degree
6 College graduate (e.g., B.A., B.S.)
7 Completed some graduate school, but no degree
8 Completed graduate school (e.g., M.S., M.D., Ph.D.)
9 Not sure
Which of the following income categories best describes your total 2002 household
income before taxes?
1 Less than $15,000
2
3
4
$15,000 to $24,999
$25,000 to $34,999
$35,000 to $49,999
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
180
5 $50,000 to $74,999
6 $75,000 to $99,999
7 $100,000 to $124,999
8 $125,000 to $149,999
9 $150,000 to $199,999
10 $200,000 to $249,999
11 $250,000 or more
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Gay men with eating disorders and food, body image, and exercise concerns: A group treatment approach

PDF

Intergenerational conflict, family functioning, and acculturation experienced by Asian American community college students

PDF

Gender differences in motivation for sexual intercourse: Implications for risky sexual behavior and substance use in a university and community sample

PDF

Assessment of racial identity and self -esteem in an Armenian American population

PDF

Family functioning, intergenerational conflict, and psychological symptomology of Asian American college students

PDF

Gender differences in symptom presentation of depression in primary care settings

PDF

Effects of test interpretation style and need for cognition on elaboration, favorability, and recall

PDF

Gender role conflict, personality, and help -seeking in adult men

PDF

Integration of science and practice: A collective case study of scientist -practitioner programs in counseling psychology

PDF

A Positive Psychology goal intervention using strengths and values to enhance goals and increase well -being

PDF

Development and validation of the Cooper Quality of Imagery Scale: A measure of vividness of sporting mental imagery

PDF

An investigation of a new diagnostic sub-type: Post traumatic stress disorder with psychotic features

PDF

Attachment style, interpersonal guilt, parental alcoholism, parental divorce and eating disordered symptomatology in college women

PDF

An investigation of the psychological well-being of unaccompanied Taiwanese minors/parachute kids in the United States

PDF

Counselor Trainees' Perceptions Of And Attitudes Toward Sexual Harassment In Relation To Feminist Beliefs

PDF

An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

PDF

Examining the differences in family functioning and meaning in life between Reform, Conservative, and Orthodox Jewish families

PDF

A daily diary approach to compare the accuracy of depressed and nondepressed participants' estimation of positive and negative mood: A test of the depressive realism hypothesis

PDF

Anger expression in men with elevated blood pressure: Relationships to assertiveness

PDF

Must everyone clean their plates? the relationship between culture, food rules and disordered eating amongst minority women

Asset Metadata

Creator Anderson, Julie Lynn (author)

Core Title Heterosexuals' attitudes toward lesbians and gay men and the attitude functions they serve: Correlates, stability and gender differences

School Graduate School

Degree Doctor of Philosophy

Degree Program Education (Counseling Psychology)

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,psychology, social

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Chung, Ruth H. (committee chair), Goodyear, Rodney (committee member), Williams, Walter L. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-475589

Unique identifier UC11334977

Identifier 3133241.pdf (filename),usctheses-c16-475589 (legacy record id)

Legacy Identifier 3133241.pdf

Dmrecord 475589

Document Type Dissertation

Rights Anderson, Julie Lynn

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA