In the rest of the logistic regressions, viewing pornographic videos was the key independent variable whereas socio-demographic factors were used as covariates and indicators of HIV-related sexual risk behaviours were considered as dependent variables.