Content uploaded by Michael Ross
Author content
All content in this area was uploaded by Michael Ross on Feb 20, 2019
Content may be subject to copyright.
Content uploaded by Michael Ross
Author content
All content in this area was uploaded by Michael Ross on Feb 20, 2019
Content may be subject to copyright.
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/14518669
Measurement and correlates of internalized homophobia: A
factor analytic study
ArticleinJournal of Clinical Psychology · January 1996
DOI: 10.1002/(SICI)1097-4679(199601)52:1<15::AID-JCLP2>3.0.CO;2-V·Source: PubMed
CITATIONS
243
READS
595
2 authors, including:
Some of the authors of this publication are also working on these related projects:
ParTy Study (HIV Risk Among Methamphetamine Using U.S. MSM) View project
Surgical management of middle ear cholesteatoma and reconstruction at the same time View project
Michael Ross
University of Minnesota Twin Cities
505 PUBLICATIONS11,255 CITATIONS
SEE PROFILE
All content following this page was uploaded by Michael Ross on 20 February 2019.
The user has requested enhancement of the downloaded file.
Body
Image
of
Eating
Disorder
Subjects
15
WATT,
R.
J.,
&
ANDREWS,
D. P.
(1981).
APE:
Adaptive probit estimation
of
psychometric functions.
Cur-
WILMUTH,
M.
E.,
LEITENBERG,
H.,
ROSEN,
J.
C.,
&
CADO, S.
(1988).
A
comparison of purging and non-
WILMWTH,
M.
E.,
LEITENBERG,
H.,
ROSEN,
J.
C.,
FONCARDO,
K.
M.,
&
GROSS,
J.
(1985).
International Journal
rent Psychology Review,
l,
205-214.
purging normal weight bulimics.
International Journal
of
Eating Disorders,
7,
825-835.
of
Eating Disorders,
4,
71-78.
MEASUREMENT AND CORRELATES OF INTERNALIZED HOMOPHOBIA:
A FACTOR ANALYTIC STUDY
MICHAEL
W.
ROSS
School
of
Public Health
University
of
Texas
B.
R.
SIMON
ROSSER
Program in Human Sexuality
University
of
Minnesota
We developed a scale to measure internalized homophobia in men who have
sex with men, which is comprised
of
items derived from theoretical and
clinical reports
of
internalized homophobia. Two hundred two men who
have sex with men and who attend “Man to Man” sexual health seminars
in a midwestern
U.S.
city completed the scale at baseline. Orthogonal factor
analysis revealed four dimensions
of
internalized homophobia: public iden-
tification as gay, perception of stigma associated with being homosexual,
social comfort with gay men, and the moral and religious acceptability
of
being gay. Factor scoring
of
these dimensions indicated that they were
associated significantly with relationship satisfaction, duration
of
longest
relationship, extent
of
attraction to men and women, proportion
of
social
time with gay people, membership
of
gay/bisexual groups,
HIV
serostatus,
and disclosure
of
sexual orientation. Internalized homophobia is measurable
and consists
of
four dimensions that are associated significantly with low
disclosure, shorter length of and satisfaction with relationships, lower degree
of
sexual attraction
to
men and higher degree
of
attraction to women, and
lower social time with gay people.
Internalized homophobia is a central construct in the symptomatology and treat-
ment of gay men and lesbians. Cabaj
(1988)
argues that it is the main dynamic in neurosis
in homosexual people, with ego-dystonic homosexuality, previously
a
diagnostic category,
being described as dissatisfaction with being homosexual and as being associated with
low self-esteem and self-hatred. Thus, the diagnostic category
of
ego-dystonic homosex-
uality included what currently is described as internalized homophobia. Cabaj suggests
that internalized homophobia will affect symptomatology, transference, and counter-
transference in psychotherapy with homosexual people. In addition, Kahn
(1991)
has
noted that theories of homosexual identity formation are based on the assumption that
The “Man to Man” seminars were funded by a grant
from
the AIDS/STDs Program
of
the Minnesota
Correspondence should be addressed
to
Dr.
Michael
Ross,
Center
for
Health Education Research and
Department of Health.
Development, School of Public Health, University of Texas, PO Box
20186,
Houston, TX
77225.
16
Journal
of
Clinical Psychology, January 1996,
Vol.
52,
No.
1
internalized homophobia is a reaction to societal homonegativism that must be resolved
for adequate psychological integration of the individual’s sexuality to occur. She argues
that this will occur by changing the meanings attached to homosexuality, such that these
will take on more positive meanings.
Cabaj distinguishes between external homophobia, the stigma associated with being
homosexual, and internalized homophobia, the reaction to that homonegativism. It is
conceived of as
a
continual variable. Coleman, Rosser, and Strapko
(1992)
note that
although self-identification as gay or lesbian indicates
a
degree of self-acceptance, it
does not leave the individual immune to the prevalent external societal homophobia.
While the term “homophobia” has the widest currency, it is a set of negative attitudes
and beliefs about homosexuality, rather than a phobia as classically understood (Herek,
1984a, 1984b;
Shields
&
Harriman,
1984).
Internalized homophobia is described as the
internalization of negative attitudes and assumptions about homosexual people (Sophie,
1987).
Malyon
(1982)
has conceptualized internalized homophobia as being both con-
scious and unconscious, which makes homosexuality ego-alien. He argues that the first
stage of psychotherapy in gay people is to identify the presence and nature of inter-
nalized homophobia.
The psychological characteristics associated with internalized homophobia have been
described
as
lower self-acceptance, lower ability to self-disclose to heterosexual and other
homosexual persons (Kahn,
1991),
low self-esteem, self-hatred, self-doubt, belief in one’s
inferiority, acceptance of popular myths about homosexuality, beliefs that others will
be rejecting on the basis of one’s sexuality, and self-imposed limits on one’s aspirations
(Cabaj,
1988).
Cabaj has listed the specific beliefs that characterize internalized homo-
phobia as being that older gay men are lonely and sad; that given the choice, gays would
elect to be heterosexual; that gays can’t have lives
as
fulfilling
as
those of their heterosexual
counterparts; and that there are social phobias about interacting in gay environments.
With regard to behavior, Coleman et al.
(1992)
add to this list that hypersexual behavior,
anonymous sex, avoidance
of
relationships, avoidance of intimacy, and substance use
can serve the function of escaping confrontation with one’s own homophobia. Further,
Gonsiorek
(1988)
reports that an additional manifestation is that gay people hold
themselves to higher standards of conduct compared with heterosexuals and that one
of
the most sensitive indicators is the way in which individuals view other members of
the homosexual community.
The attributed impact of internalized homophobia is broad. Both Remafedi
(1987)
and Gonsiorek
(1988)
note that it is
a
major problem for adolescents, and Dunne
(1987)
reports that it is
a
significant variable in disclosure in gay fathers. Cabaj
(1989)
sees it as
a significant variable in substance abuse in gay and bisexual men. Thus, while there is
general agreement about the theoretical importance of internalized homophobia, there has
been little operationalization of the construct. Although Kahn
(1991)
reports on an
18-
item questionnaire based on myths related to homosexuality, no data on the content or
psychometrics
of
the scale are provided. Kahn found that internalized homophobia was
related significantly and negatively to feminist orientation and discomfort in disclosing
sexual orientation in lesbians. Nicholson and Long
(1990)
studied
89
HIV seropositive
Canadian gay men and used the Nungesser Homosexual Attitudes Inventory (NHAI) as
a measure of homophobic prejudice. They found that there was a relationship between
internalized homophobia and self-blame-related coping and avoidance coping styles. The
NHAI is based on three subscales, which comprise attitudes toward one’s own homosex-
uality, toward other homosexuals, and toward disclosure, which had alpha coefficients
of
.68
to
.94.
The
34
items were based on self-reports of homophobic prejudice and
have not been subject to factor analysis, although they were chosen from a larger pool
of items by item analysis. Because of the lack of a scale based on both the theoretical and
clinical components of internalized homophobia and the need to assess concurrent valid-
ity
of
any scale, we determined to develop a scale to measure internalized homophobia
Internalized Homophobia
17
and to measure the relationship of internalized homophobia to sexual attraction, rela-
tionship satisfaction and length, social interaction with gay people, and disclosure.
METHOD
Subjects and Procedure
Participants in the study were all males over
18
years of age who responded to an
opportunity to attend one of four “Man-to-Man Sexual Health Seminars”
in
1993.
These
seminars were advertised as an opportunity for “men who have sex with, or are attracted
to, other men” to explore issues of “intimacy between men.” Because the term “homosex-
ually active” was seen as potentially threatening to some interested participants, “men
who are attracted to other men” was added to relieve participants
of
the need
to
iden-
tify themselves by their behavior. Advertisements noted that the seminars were free (being
paid for by a grant from the Minnesota Department of Health), included a dinner, and
were conducted by licensed psychologists who specialize in human sexuality. The first
and third seminars were conducted on a Friday-Saturday and the second and fourth
on a Saturday-Sunday, all at the University of Minnesota. Of the 202 seminar par-
ticipants,
36
attended the first seminar,
54
the second,
57
the third, and
55
the fourth.
Initial recruitment targeted homosexually active men in difficulty by soliciting referrals
from Program in Human Sexuality clients; from other therapists, medical practitioners,
and HIV clinics; and from service agencies that target individuals at risk for HIV (such
as the Minnesota AIDS Project).
As
a consequence, the study sample could be expected
to be skewed toward those who are experiencing psychological and psychosexual prob-
lems:
21
Yo
of the sample was recruited through the above sources. Advertising in local
gay and community magazines (GAZE, Equal Times), pamphleting and posters at gay
events (the Gay, Lesbian, and Transgender Pride Festival), and gay venues and churches
recruited the remainder of the participants.
Interested participants could register by providing a name (real or false), and con-
firmation of reservation was sent to each participant who registered in advance. In ad-
dition, registrations were accepted at the time
of
the workshop on an as-available basis.
Registration was conditional on the person signing a consent form that indicated that
he was over
18
years of age, knew he would be exposed to sexually explicit material,
and knew, if he chose to be involved in the research study, that he could refuse
to
answer
any question and could withdraw from the study at any time. When participants had
registered and received a name tag, they were to participate separately in the research/
evaluation component of the seminar. All participants signed the research form, and
none refused any part
of
the questionnaires. The baseline questionnaires took about
20
minutes to complete, the post-seminar,
15
minutes. Post-seminar questionnaires were
handed out in the seminar
30
minutes before its conclusion.
Data for the “Man-to-Man” seminar were collected at baseline, post-seminar, and
at
2
months follow-up. To assess behavioral change, data on sexual and drug behavior
in the previous 2 months were collected at baseline and 2-month follow-up: The data
reported on here are the baseline data. Demographic and relationship data were col-
lected at baseline, along with responses to research questions that assessed knowledge
and attitudes with regard to comfort with sexuality, safer sex, condom use, assertiveness,
mental health, and self-esteem.
Data analysis consisted of subjecting the items
of
the internalized homophobia scale
at baseline to factor analysis (principal component analysis followed by varimax rota-
tion to simple structure). Subscales were computed by multiplying the item score by
the factor loading after recoding those items scored in
a
reverse direction. Intercorrela-
tions were calculated using Pearson product-moment correlation coefficients, and in-
ternal reliabilities by Cronbach’s alpha coefficient, while comparisons of group means
were carried out using t-test.
18
Journal
of
Clinical Psychology, January
1996,
Vol.
52,
No.
I
RESULTS
Completed data for baseline and post-seminar responses were available for
184
men.
The majority were college graduates
(33.7%)
or had
a
graduate degree or professional
qualification
(31.7%).
Mean age was
37.0
years,
SD
=
9.3
years. The factor analysis
of the baseline data produced seven factors with eigenvalues
>
1.
However, several of
these were monofactors
or
factors with fewer than three unique items or with loadings
>
.30,
so
the data were re-rotated on the basis of a scree test and four factors extracted,
which accounted for
45.1’70
of the variance. These four factors (listed in Table
1)
pro-
duced readily interpretable dimensions
of
(1)
public identification as gay;
(2)
perception
of
stigma associated with being gay;
(3)
social comfort with gay men; and
(4)
moral
and religious acceptability of being gay. The scales computed from these dimensions
had internal reliabilities (coefficient alpha) of
.85,
.69,
.64,
and
.62,
respectively.
The subscales’ correlations with the measures of validity are presented in Table
2.
There were
a
number of significant associations; public identification as being gay
Table
1
Factor Structure
of
the Internalized Homophobia Scale
Factor
1:
Public Identification as Gay
23.
11.
25.
12.
21.
10.
19.
8.
1.
3.
I
am not worried about anyone finding out that
I
am gay.
I
feel comfortable discussing homosexuality in
a
public setting.
Even if
I
could change my sexual orientation,
I
wouldn’t.
It is important to me to control who knows about my homosexuality.
I
feel comfortable about being homosexual.
I
feel comfortable about being seen in public with an obviously gay person.
I
would prefer to be more heterosexual.
I
don’t like thinking about my homosexuality.
Obviously effeminate homosexual men make me feel uncomfortable.
It would not be easier
in
life to be heterosexual.
(23.8%
of variance)
Factor
2:
Perception of Stigma Associated with Being Gay
17.
I
worry about becoming
old
and gay.
18.
I
worry about becoming unattractive.
15.
Society still punishes people for being gay.
13.
Most people have negative reactions to homosexuality.
24.
Discrimination against gay people is still common.
20.
Most people don’t discriminate against homosexuals.
(9.1
Qo
of
variance)
Factor
3:
Social Comfort with Gay Men
6.
I
feel comfortable in gay bars.
4.
Most of my friends are homosexual.
5.
I
do not feel confident about making an advance to another man.
9.
When
I
think about other homosexual men,
I
think of negative situations.
7.
Social situations with gay men make me feel uncomfortable.
2.
I
prefer to have anonymous sexual partners.
(6.4%
of variance)
Factor
4:
Moral and Religious Acceptability of Being Gay
14.
Homosexuality is not against the will of God.
22.
Homosexuality is morally acceptable.
26.
Homosexuality is as natural as heterosexuality.
16.
I
object if an anti-gay joke is told in my presence.
(5.8%
of variance)
.71
.13
.69
.61
.60
-
.59
-
.59
-
.41
.36
-
.67
.76
.71
.59
.44
.44
-
.41
.62
.59
-
.55
-
.54
-
.50
.48
.65
.60
.56
.42
Internalized Homophobia
19
Table
2
Variables Associated with Internalized Homophobia Factors
~ ~~ ~
Public Stigma Comfort Acceptability
Correlations
Duration longest relationship
Extent
of
attraction to men
Extent of attraction to women
Relationship satisfaction
Proportion of social time with gays
Openly gay/bisexual in personal life
Openly gaylbisexual at work
Number known with HIV/AIDS
HIV seropositive
HIV seronegative
Belong gay/bi group
Do not belong gay/bi group
Means on scales
-
.23*
-
.26*
.35**
-
.26*
.39**
-
.57**
-
.64**
-
.35**
13.45
18.18**
16.27
20.48.
.I6
-
.05
.01
.25*
-
.04
.07
.13
-
.03
10.57
10.27
10.09
10.21
-
.01
-
.32**
.27**
-
.29**
.53**
-
.36**
-
.27**
-
.30**
8.42
11.93**
1
I
.05
12.35*
-
.25*
-
.17*
.24*
-
.08
.09
-
.19*
-
.08
-
.14
3.45
4.15
3.77
4.51*
was associated with all eight measures, and social comfort with other gay men was
associated with seven of the eight. Moral and religious acceptability of being gay was
associated with four of eight measures, while perception of stigma associated with being
gay was only associated with relationship satisfaction. HIV seropositive respondents
(12/153)
were significantly lower on r-test on public identification
of
being gay and on
social comfort with other gay men. There was
a
similar pattern
on
belonging to a gay
or bisexual group, with the addition of moral and religious acceptability of being
gay.
Factor intercorrelations were factor
1
with factor
2,
-
.35
(p
<
.OOOl),
with factor
3,
.45
(p
<
.OOOl),
with factor
4,
.35
(p
<
.OOOl);
factor
2
with factor
3,
-
.27
@
<
.001),
with factor
4,
-
.03
(ns);
and factor
3
with factor
4,
.24
(p
<
.002).
DISCUSSION
These data must be interpreted with the caveat that they are based on samples of
gay men who have volunteered for seminars on sexual health, and as such the sample
is biased toward those who have accepted to some degree their sexual orientation. Fur-
ther, the sample is of moderate size and may not be typical
of
gay men outside of the
midwestern region of the United States. It appears to be skewed toward the better
educated and mature. These cautions aside, it is apparent that internalized homophobia
is measurable and that empirically it is categorized by four distinct dimensions, all of
which are moderately intercorrelated.
The results
of
the factor analysis presented in Table
1
suggest that these four dimen-
sions measure
(1)
concern about public identification about being gay;
(2)
concern about
the stigma associated with being considered unattractive (in the gay world) and the stigma
associated with being gay outside the gay world. (These first two items, which measure
the former, were a separate factor in the five-factor solution.);
(3)
social comfort with
other gay men; and
(4)
the moral and religious acceptability of being gay. Theoretical
conceptions of internalized homophobia stress the internalization of anti-homosexual
sentiment as resulting in concern about being identified and discomfort with one’s iden-
tification, and if a person does exhibit internalized homophobia, then discomfort with
other gay men (and preference for anonymous sexual encounters) would be anticipated.
This is consistent with perception of stigma associated with being gay, identified by
Ross
20
Journal
of
Clinical Psychology, January
1996,
Vol.
52,
No.
1
(1983,
1985)
as being associated with heterosexual marriage of homosexual men, as being
unrelated to the
actual
level of societal discrimination experienced, but significantly
associated with the
anticipated
discrimination. These data suggested that it was the
perception and anticipation
of
negative response to sexual orientation, rather than the
actual response, which were associated with discomfort and attempting to downplay
or
hide orientation, and the present data confirm that the perception of stigma associated
with being gay is a component of internalized homophobia. The final dimension ex-
tracted indicates that a moral or religious concern about the acceptability of being gay
is, independent of societal stigma,
a
factor in internalized homophobia. These four
dimensions of internalized homophobia are consistent with the previous literature on
internalized homophobia and, more importantly, indicate that internalized homophobia
in homosexually active men can both be measured empirically and exhibit internal
consistency.
The four subscales that make up the Internalized homophobia scale also exhibit
significant concurrent validity when compared with the criterion measures in Table
2.
It is clear that the two subscales with the most consistent reponse are those that measure
concern with public identification as being gay, and comfort in gay social contexts. To
a lesser extent, moral and religious concerns about homosexuality also are associated
with a majority of these measures. Because the variables selected for measurement of
concurrent validity are all theoretically or clinically identified outcomes
of
internalized
homophobia, the consistency of these correlations with the Internalized Homophobia
Scale subscale scores is further evidence that
this
scale does tap the clinical construct
of internalized homophobia. The association of high Public identification, and Social
comfort with gay people, is probably an epidemiological function of the fact that HIV
seroprevalence in Minnesota is predominantly in the population identified as gay.
These data do suggest that the clinical construct of internalized homophobia is
measurable and psychometrically has both internal reliability and concurrent validity.
Nevertheless, these data should be replicated using large and more divergent popula-
tions, and, if possible, in samples in which internalized homophobia has been diagnosed
independently as
a
further clinical validation. The present data suggest that internalized
homophobia is
a
measurable construct and that its major dimensions are Public iden-
tification as being gay, Perception
of
stigma associated with being gay, Moral and
religious acceptability of being gay, and Social comfort with gay men. These dimen-
sions are associated with variables that have been identified clinically as associated with
internalized homophobia.
REFERENCES
CARAJ,
R.
P. (1988). Homosexuality and neurosis: Considerations for psychotherapy. In
M.
W.
Ross
(Ed.),
7’he treatment of homosexuals with mental health disorders
(pp. 13-23). New York: Harrington Park Press.
COLEMAN,
E.,
ROSSER,
B.
R.
S.,
8r
STRAPKO,
N. (1992). Sexual and intimacy dysfunction among homosexual
men and women.
Psychiatric Medicine,
10,
257-271.
DUNNE,
E.
J.
(1987). Helping gay fathers come out to their children.
Journal of Homosexuality,
14(1
&
GONSIOREK,
J.
C.
(1988). Mental health issues
of
gay and lesbian adolescents.
Journal of Adolescent Health
Care,
9,
114-122.
HEREK,
G.
M. (1984a). Beyond “homophobia”:
A
social psychological perspective on attitudes toward lesbians
and gay men.
Journal
of
Homosexuality,
IO(1
&
2). 1-21.
HEREK,
G.
M.
(1984b). Attitudes toward lesbians and gay men:
A
factor-analytic study.
JournalofHomosex-
uality,
1q1
&
2), 39-51.
2), 213-222.
KAHN,
M.
J.
(1
991). Factors affecting the coming out process for lesbians.
Journal
of
Homosexuality,
21,
47-70.
MALYON,
A.
(1982). Psychotherapeutic implications of internalized homophobia in gay men.
Journal
of
Homosexuality,
7(2
&
3), 59-69.
Internalized Homophobia
21
NICHOLSON,
W.
D.,
&
LONG,
B.
C.
(1990).
REMAFEDI,
G.
(1987).
Homosexual youth:
A
challenge to contemporary society.
JAMA, 258,
222-225.
Ross,
M.
W.
(1983).
The married
homosexual
man:
A
psychologicalstudy.
London: Routledge
&
Kegan Paul.
Ross,
M.
W.
(1985).
Actual and anticipated societal reaction to homosexuality and adjustment in two societies.
Journal
of
Sex Research, 21,
40-55.
SHIELDS,
S.
A.,
&
HA”,
R.
E.
(1984).
Fear
of
homosexuality: Cardiac responses of low and high
homonegative men.
Journal
of
Homosexuality,
lO(1
&
2), 53-67.
SOPHIE,
J.
(1987).
Internalized homophobia and lesbian identity.
Journalof Homosexuality,
14(1&
2), 53-65
Self-esteem, social support, internalized homophobia and coping
strategies
of
HIV
+
gay men.
Journal
of
Consulting and Clinical Psychology,
58,
873-876.
SOCIAL MALADJUSTMENT INDICATORS
IN PTSD PATIENTS FAMILIES
OF
ORIGIN
CHARLES
G.
WATSON, PATRICIA
E.
D.
ANDERSON
AND
LEE
P.
GEARHART
Department
of
Veterans Affairs Medical Center
St. Cloud, Minnesota
We compared the frequencies with which
PTSD
patients, psychiatric con-
trols, and hospital employee controls reported that their fathers, mothers,
and oldest siblings
of
each sex had been incarcerated
or
had received
psychiatric/psychological
treatment. We also compared estimates
of
the
number of psychiatric hospitalizations, incarcerations, courses
of
outpatient
treatment, treatment sessions, and days of institutionalization undergone
by the relatives. Only a chance number of significant differences appeared,
which suggests that general psychosocial maladjustment in one’s family
of
origin does
not
appear
to
increase trauma survivors’ risk for
PTSD.
A
thorough knowledge
of
the relationship between a psychiatric disorder and
psychopathology in its victims’ families of origin is important to a good understanding
of its etiology. Two recent studies (True et al.
1993;
Watson, Anderson,
&
Gearhart,
unpublished) suggest that a moderately genetic predisposition to post-traumatic stress
disorder (PTSD) is passed along within families. This raises a question as to whether
the presence
of
other psychiatric disorders or psychosocial maladjustment problems in
a trauma survivor’s family of origin increases hidher risk for PTSD. The literature on
this question is inconsistent. Kulka et al.
(1988,
p.
F-10)
reported that a family history
of mental illness increased risk for PTSD in Vietnam war veterans, but Speed, Engdahl,
Schwartz, and Eberly
(1989)
found no relationship between a family history of psychiatric
disorder and PTSD in their World War I1 prisoners. Emery, Emery, Shama, Quiana,
and
Jassani
(1991)
reported a high incidence of alcoholism in the parents of their PTSDs,
but their results conflict with Davidson, Smith and Kudler’s
(1989)
on this point. Davidson
et al. also found that their PTSD patients’ families of origin had exaggerated levels of
This research was supported by the Department of Veterans Affairs Medical Research Service. Requests
for reprints should be addressed to Charles
G.
Watson, Research Service, Department
of
Veterans Affairs
Medical Center,
4801
8th St.
N,
St. Cloud, MN
56303.
View publication statsView publication stats
