Multicultural Considerations for the Psychometrics of the Brief Pornography Screen.

Abstract

Problematic pornography use (PPU) is a burgeoning area of clinical interest. The Brief Pornography Screen (BPS) is a new PPU measure. The BPS has not been psychometrically studied within specific race/cultural groups. We sought to broaden the PPU literature by examining the confirmatory factor analysis (CFA) fit, measurement invariance, and structural invariance of the BPS across eight identity groups. Participants came from a survey administered at three U.S. universities. In total, n = 2,475 participants were analyzed, with the following identity group breakdowns: Asian American/Pacific Islander (AAPI) male = 268, AAPI female = 303, Black male = 101, Black female = 189, Latin male = 208, Latin female = 372, White male = 432, and White female = 602. BPS CFA fit was good across all groups. Measurement invariance analyses suggested metric, but not scalar invariance across all groups. We then split participants by sex assignment, full residual invariance was evident across groups for male participants and partial residual invariance was evident for female participants. Structural invariance analyses indicated anxiety as a weak positive BPS correlate in AAPI, Latin, and White male participants (β's = 0.25-0.27), but not meaningfully related in the other groups. Pornography viewing frequency was positively correlated with BPS scores across most groups with a wide range (β's = 0.29-0.52), except for Black male participants (β = 0.15). Our results suggest that the BPS is an appropriate PPU measure across the tested identity groups. While between-group measurement is relatively accurate within sex assignment groups, correlates differed in strength, meaning different variables likely predict PPU for different groups of people.