TPC-Journal-V1-Issue3

168 The Professional Counselor \ Volume 1, Issue 1 Materials Levels of Outness . The questions used in this study that assessed level of outness of participants were adapted with permission from a survey developed by Rankin (2003). These questions are part of a larger campus climate survey that is used nationally to assess campus climate for community members. This set of questions showed high internal consistency (α = .80). Using a varimax rotation, items load as one factor and account for 65% of the variance. Internalized Homophobia Scale (IHP; Martin & Dean, 1987). The Internalized Homophobia Scale is a 9-item measure adapted for self-report. Previous research has indicated that the self-administered version of the IHP scale has acceptable internal consistency and correlated as expected with relevant measures (Herek & Glunt, 1995). Items are administered with a 5-point response scale, ranging from disagree strongly to agree strongly . Reliability coefficients for this scale are typically higher for men (α = .83) then women (α = .71). Procedure An online survey was created through PsychData, a Web-based company that conducts Internet-based research in the social sciences. Participants were invited through email advertisements to general and multicultural LGBT list-servs. Personal networks also were utilized. Participants were made aware of the intentions of the survey and the topics they would encounter. Participation was anonymous and all respondents reviewed and gave informed consent before initiating the study. (Appropriate IRB approval was obtained). Results A set of regression procedures were conducted to examine how “coming out” to friends, family, and colleagues predicted scores on the Internalized Homophobia Scale (IHP). Table 2 shows the complete regression summaries. Model 1 regressed demographic variables: age, self-identified gender, ethnicity, education level and reported income. This model accounted for 4% of the overall variance with education level contributing most to the prediction of IHP scores (β = -.15, p < .05). This model was used in each of the following hierarchical models as a first step to control for demographics to examine the prediction value of each “coming out” variable. Model 2 emerged as a significant model. Coming “out to friends” was added as a second step to the model and accounted for an additional 11% of the variance after controlling for demographics. According to the standardized regression coefficients, this variable provided a significant contribution to the prediction of IHP scores (see Table 2). Model 3 did not emerge as a significant model. By including coming “out to nuclear family” only an addition 1% of the total variance was accounted for. Within the third model, coming “out to friends” remained the largest contributor to the prediction of internalized homophobia (β = -.28, p < .05). Model 4 and Model 5 emerged as significant models. In Model 4, the second step added coming “out to extended family” and contributed 2% to the total variance. Standardized regression coefficients indicated coming “out to friends” and “out to extended family members” were the strongest contributors to the model. By adding “out to colleagues” to the second step of the fifth model, an additional 2% of the variance was accounted for beyond demographics. “Out to colleagues” contributed the most to the predication of internalized homophobia beyond demographic differences, coming out to friends, coming out to nuclear family, and coming out to extended family (β = -.18, p < .05). Model 5 accounted for almost 19% of the total variance.