Does questionnaire format impact reported partner violence rates?: An experimental study

Researchers assess partner violence using numerous formats, but whether questionnaire format affects obtained rates has rarely been examined. We compared paper-and-pencil versus computer administrations, and standard Revised Conflict Tactics Scales (CTS2) categories versus dichotomous (yes/no) response categories, in a 2 x 2 experiment with 160 undergraduate students. Differential validity was also examined for gender. Results provide some support for the use of all formats, although sexual coercion rates showed more measurement variability than reports of psychological aggression, physical assault, or injury. Sexual coercion effects also differed by gender. The standard CTS2 categories resulted in more disclosures than dichotomous categories for some variables. Differences between computer and paper-and-pencil formats were generally minimal. More attention to the effects of questionnaire design on disclosures of violence is urgently needed.