To determine whether gay-sensitive HIV
instruction was associated with reduced sexual
risk behaviors, we compared the responses of
GLB adolescents in schools with and without
gay-sensitive HIV instruction, using the analytic
procedures described above. Finally, we
compared GLB and heterosexual youths in
schools with and without gay-sensitive HIV
instruction, using 2 (GLB vs heterosexual
youths) 2 (no/minimal/low vs moderate/
high gay-sensitive-instruction schools) ANCOVAs
and parallel logistic regression analy