Effects of response mode, response format, perceived response style, sex, and anonymity on raw scale scores and extreme response tendencies for two distinct measures of affective behavior were investigated. The two measures were a self-acceptance scale developed by J.R. Phillips and the Machiavellianism IV scale. Where (response mode) and how (response format) the respondents were to report their responses were manipulated in seven different response forms. Two hundred twelve college age and adult subjects participated in the investigation. Findings indicated statistically significant effects for response mode, format and style. Results suggest that the effects of response form are a function of both subject and instrument characteristics. (Author/MV)