McDermott, Tyndall, and Lichtenberg

selected variables in addition to the scant information available to mental health professionals.

Participants

METHOD

Participants in this study were homosexually oriented men and women from the midwest. A total of 47 men and 36 women responded to the survey. Of the sample, 89% was Caucasian, with the remainder including Native American, Black, and Hispanic individuals. Respondents ranged in age from 18 to 68 years of age, with a mean age of 28 (SD=7.6). A majority of the sample (98%) had completed high school or had pursued higher education, with 35% having received their B.A., 17% having obtained their M.A., and 2% having completed their Ph.D. Approximately 37% of the participants reported an income under $10,000 with another 45% reporting an annual income between $10,000 and $19,000. About 19% of the sample indicated that its annual income was over $20,000. Roughly onethird of the participants described themselves as students, while one-third responded that they were in professional occupations. The remainder primarily fell into white-collar, skilled-trade, and unskilled labor categories. No significant differences were found between sexes on any of these demographic variables.

Procedures

After obtaining approval from local gay and lesbian organizations, the investigators attended gay and lesbian events on the university campus of a large midwestern university and in the local community of that university and asked individuals attending those functions (workshops, dances, and potluck gatherings) to participate in research concerned with gay and lesbian preferences for counselors. A majority of participants filled out the four-page questionnaire at the time their participation was requested, although some returned surveys by mail.

Instruments

Four instruments were used in this study.

Participant Questionnaire. The first instrument was a questionnaire concerned with obtaining demographic data (e.g., age, ethnic group, education, income, occupation, and status, whether partnered or single). Other background information was also solicited by the questionnaire. Included were items exploring how many and which groups (friends, parents, siblings, coworkers, and clergy) the respondents had informed of their sexual orientation, an estimate of the percentage of people to whom respondents were "out," and three items using a 5-point Likert-type rating scale (1 = not at all, 5=very) asking about the significance of emotional involvement with sexual partners, the degree of satisfaction with sexual contacts, and how sexually active respondents were. In addition, one question inquired about the participants' preference for the sexual orientation (heterosexual, bisexual, gay/lesbian, or no preference) of a counselor should they choose to see one; another question asked respondents who had received counseling to indicate the sexual orientation of their counselor, if it was known. Finally, three items, using the same 5-point Likerttype rating scale, were included for participants to indicate their satisfaction with previous counselors.

Kinsey Scale. The second instrument was the Kinsey Scale (Kinsey et al., 1953), a single-item, 6-point scale, which asked respondents to choose the category that best described them in

222

terms of their preferences for sexual involvement with members of the same sex, opposite sex, or both. The scale ranged from 0 (indicating an exclusive preference for sexual relations with members of the opposite sex and no interest in sexual relations with the same sex) to 6 (indicating an exclusive preference for sexual relations with members of the same sex and no interest in sexual relations with members of the opposite sex).

Counseling Concerns Scale. The third instrument was a list of 14 concerns that people commonly discuss with a counselor. Participants were asked to indicate on a 5-point Likert-type scale their comfort level in discussing each item with a counselor whose sexual orientation they did not know (1 = very uncomfortable, 5= very comfortable). Items were selected to represent issues commonly brought to a university counseling center but seemed, to the authors, to be either central or peripheral to an individual's sexual identity. With the exception of one item ("Stressful issue of school or work overload"), the logical "twofactor structure" of the questionnaire (i.e., central versus peripheral concern) was confirmed by a Varimax factor analysis. In all analyses using this instrument, items designated as central or peripheral were those confirmed by the factor analysis.

Index of Homophobia. The final instrument was the 25-item Index of Homophobia (IHP) (Hudson & Ricketts, 1977), which uses a 5-point rating scale (1 = strongly agree, 5= strongly disagree) to assess the way people feel about working and associating with homosexuals.

RESULTS

Overall, 49% of the total sample indicated a preference for a gay or lesbian counselor, while 39% reported that they did not think that their counselor's sexual orientation made a difference. Approximately 10% of the participants indicated a preference for a bisexual counselor, and only one individual stated a clear preference for a heterosexual therapist. A chi-square analysis of participants' preference for counselor's sexual orientation (heterosexual, bisexual, gay/lesbian, or doesn't make a difference) by sex revealed no significant difference, x2=7.13, df=3,p>.05.

An interesting finding was that 82% of the sample had been to at least one counselor. Of those individuals having experience with therapy, 46% responded that their counselor was heterosexual; 27% had seen a gay or lesbian therapist; 5% had visited a bisexual counselor; and 22% did not know their therapist's sexual orientation.

The 14 items that assessed participants' comfort in discussing various topics with a counselor of unknown sexual orientation were summed and averaged to yield a total comfort score for each participant. These total comfort scores were then analyzed in a stepwise multiple regression that revealed that the Index of Homophobia (IHP) was the best predictor of comfort in discussing various topics with a counselor, F(1,57) = 8.79, p<.01. Table 1 summarizes the means and standard deviations of the participants on each of the scales. Because the total comfort scores combined comfort ratings for the two concern factors, separate Pearson correlations were computed between participants' IHP scores and their composite ratings on the counseling concerns that were (a) central to or (b) peripheral to their sexual identity. The composite ratings for the two sets of concerns (as defined by the factor analysis) were calculated for each participant. Each participant's ratings for concerns #2, 3, 6, 7, 8, 10, 11, 12, 13, and 14 were summed to yield that participant's composite rating for counseling concerns, which were central to her or his sexual identity. Concerns #1, 4, 5 and

JOURNAL OF COUNSELING & DEVELOPMENT SEPTEMBER/OCTOBER 1989 VOL. 68

TABLE 1

Counselor Preference Among Gays and Lesbians

Means and Standard Deviations of Participants' Comfort Ratings to Various Counseling Concerns, Index of Homophobia, and Kinsey Scale

Concerns

1. Issues of career choice

2. Break up with a lover

3. Decision to come out to my parents

4. Roommate problems

5. My academic program

6. How I feel about my sexuality

Men1

Women2

All

4.57

1.76

4.53

1.00

4.56

1.10

3.57

1.44

3.83

1.16

3.69

1.33

3.70

1.56

3.69

1.43

3.70

1.50

4.32

1.16

4.28

1.19

4.30

1.17

4.36

1.33

4.36

1.18

4.36

1.26

4.28

1.16

3.86

1.25

4.10

1.21

7. Stressful issues of school or work overload

4.53

1.04

4.50

1.03

4.52

1.03

8. Coming out to my friends

4.19

1.19

3.94

1.09

4.08

1.15

9. Making job applications

4.34

1.32

4.33

1.12

4.34

1.23

10. Dating concerns

4.11

1.27

3.75

1.13

3.95

1.22

11. Health concerns (specifically, sexually transmitted diseases) 12. Issues of personal growth

3.96

1.38

3.92

1.11

3.94

1.26

4.43

1.04

4.22

0.96

4.34

1.00

13. Relationship with my family

4.26

1.09

3.86

1.44

4.08

1.26

3.60

1.61

4.17

1.11

3.84

1.44

40.62

9.90

39.75

8.93

40.24

9.45

17.60

4.38

17.50

4.14

17.55

4.25

40.92

15.87

37.89

15.25

39.60

15.58

5.38

1.13

5.42

0.84

5.40

1.01

14. Relationship with my lover

Central concerns3 Peripheral concerns4

Index of Homophobia Kinsey Scale

1N = 47.

2N = 36.

3Central concerns are those numbered 2, 3, 6, 7, 8, 10, 11, 12, 13, and 14. 4Peripheral concerns are those numbered 1, 4, 5, and 9.

9 were summed to yield the individual's composite for concerns, which were more peripheral to sexual identity.

For total sample, the correlation between the IHP and the composite score for central concerns was r=.31, p<.01, indicating that the more homophobic the participants, the more uncomfortable it was for them to discuss concerns central to their sexual identity. The relationship between the IHP and the composite score for peripheral concerns was virtually nonexistent, r = .03, p>.05. A similar pattern of correlations was found for the gay men in the sample: The correlation between their IHP score and their composite score on the central concerns was r=.37, p<.01; their correlation between the IHP and the peripheral concerns was r= .02, p>.05. For the women in the sample, the correlations between the IHP and the composite scores for central and peripheral concerns were not statistically significant, r.23 and -.09.

A separate stepwise multiple regression analysis was also conducted using the IHP as the criterion variable. This revealed that participants' estimate of the percentage of people to whom they had revealed their sexual orientation (i.e., were "out") was the strongest predictor variable, F(1,57)=12.09, p<.001. The Pearson correlation between the estimated percentage of people to whom respondents were "out" and the IHP scores was also significant, r=.44, p<.001 (i.e., the more homophobic the individual, the smaller the percentage of persons to whom they were "out.").

Separate one-way analyses of variance (ANOVAs) for sex were performed on each of the 14 comfort rating items, with only one item showing a significant difference between men and women. On item 6, "How I feel about my sexuality," men reported significantly more comfort talking with a counselor of unknown sexual orientation than did women, F(1,80) = 4.31, p<.05.

A main effect for sex was also discovered in the ANOVA for whether respondents would describe themselves as currently

"partnered," F(1,81)=4.04, p<.05. Significantly more women (57%) described themselves as partnered than did men (29%). Likewise, an ANOVA revealed that women described themselves as more sexually active than did men, F(1,81)=9.02, p<.01. No significant differences were found between men and women in terms of the significance of having an emotional involvement with sexual partners nor how satisfying sexual contacts usually were.

In terms of the percentage of people to whom participants were "out," the range was from 1% to 100% and the mean was 53% (SD=31). Of the sample, 96% reported that they were "out" to friends; 54% indicated that they were "out" to parents and siblings; 65% said that they were "out" to coworkers/fellow students; 43% reported being "out" to their employer or professors, and 28% indicated they were "out" to relatives other than immediate family. The only significant difference between sexes in terms of groups to whom they were "out" was for clergy. One-way analysis of variance revealed a main effect for sex, F(1,81)=3.90, p=.05, with more men reporting having revealed their sexual orientation to clergy (49%) than did women (28%).

Table 2 summarizes the self-reported sexual preferences of the participants as reported on the Kinsey Scale. No significant differences were found between sexes on this scale.

DISCUSSION

The primary findings of this study revealed that nearly half of the participants preferred either a gay or lesbian counselor. A surprising number of respondents (39%), however, reported that they thought that a counselor's sexual orientation did not make a difference. This latter finding provided evidence of an accepting attitude on the part of many gay men and lesbians in their view of counseling that may be unwise given the previous findings of McDermott and Stadler (1988), which suggested that

JOURNAL OF COUNSELING & DEVELOPMENT SEPTEMBER/OCTOBER 1989 VOL. 68