Table 2.
Bivariate and Multiple Regression Analyses Predicting Internalized Homonegativity Among Participants in 35 Social Networksa
| Univariate Regressions | Multiple Regression | |||
|---|---|---|---|---|
| Variable Domain/Predictor: | Coefficient (SE) | p | Coefficient (SE) | p |
| Sociodemographic Variables: | ||||
| Age (in years) | 0.09 (0.05) | .063 | ||
| Hispanic ethnicity | −1.42 (1.47) | .334 | ||
| Working full-time | 0.46 (0.82) | .569 | ||
| Annual income [ref: $0 - $10K]-- | ||||
| $10K - $30K | −1.74 (0.84) | .039 | ||
| >$30K | 0.83 (1.16) | .475 | 2.61 (0.94) | .006 |
| Attends school | 0.08 (0.90) | .930 | ||
| Education level [ref: Some college+]-- | ||||
| < Complete high school | 0.56 (1.09) | .607 | ||
| High school grad or GED | −0.65 (0.83) | .431 | ||
| Has an unstable housing situation | 2.38 (1.29) | .066 | ||
| Bisexual or heterosexual orientation | 6.93 (0.83) | <.001 | 4.43 (0.79) | <.001 |
| Has a main male partner | −1.57 (0.77) | .041 | ||
| HIV Testing Variables: | ||||
| Never tested or tested over a year ago | 1.30 (0.79) | .100 | 1.17 (0.65) | .071 |
| Test result was HIV-positive | 0.05 (0.89) | .958 | ||
| Psychosocial Contextual Scales: | ||||
| Religious/church involvement | 0.33 (0.07) | <.001 | 0.34 (0.06) | <.001 |
| Masculinity | 0.47 (0.09) | <.001 | 0.17 (0.08) | .039 |
| Gay community acculturation | −0.44 (0.06) | <.001 | −0.27 (0.06) | <.001 |
| Resilience | −0.43 (0.07) | <.001 | −0.35 (0.07) | <.001 |
| AIDS conspiracy beliefs | 0.29 (0.05) | <.001 | 0.20 (0.04) | <.001 |
|
Factors Related to Last Anal Intercourse
with a Male Partner: |
||||
| Occurred more than 6 months ago | 1.50 (1.07) | .163 | ||
| Casual hookup or commercial partner | 1.54 (0.89) | .084 | ||
| Met partner online | −1.40 (0.85) | .099 | ||
| Did not tell partner your HIV status | 0.23 (0.94) | .808 | ||
| Partner disclosed he was HIV+ or refused to tell his status |
0.82 (0.80) | .305 | ||
| Condom not used at last AI | −0.68 (0.77) | .375 | ||
| Used drug or alcohol at last AI | 1.14 (0.79) | .148 | ||
| Was high or buzzed at last AI | 1.25 (0.83) | .132 | ||
| Substance Use in Past 30 Days: | ||||
| Drank any alcohol | 0.90 (1.08) | .405 | ||
| Number of days drank alcohol | 0.05 (0.06) | .418 | ||
| Smoked marijuana | 1.25 (0.77) | .104 | 1.14 (0.66) | .086 |
| Number days smoked marijuana | 0.03 (0.03) | .349 | ||
| Used any street drug (excluding marijuana) | 2.52 (0.89) | .005 | 1.60 (0.77) | .039 |
| Number days used street drugs | 0.08 (0.07) | .317 | ||
a
A backwards stepwise algorithm was used to select covariates for the final mixed regression model from all predictors that achieved at least a p-value < .15 in the univariate analysis. Data were presented for all covariates in the multiple regression that achieved a p-value < .10.