≈ HIV testing stigma was dichotomized at the mean due to the distribution of responses. Whereas HIV coping self-efficacy was used as a continuous scale due to the normal distribution of responses. * Factors significant at 0.10 or less in simple logistic regression and theoretical factors where included in the final model. The final adjusted model was controlled for pregnancy. Prior HIV testing history was dropped from the final model due to co-linearity. ‡ P value for the final model using multiple logistic regressions. Variables at p value <0.05 included in the final model. £ Range of self-reported number of partners among youth who reported being not sexually active (0-9). |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Table 4: Factors Associated with HIV Testing in Never Sexually Active Adolescents (N=18, 22%). |