Impression management pressures on racial attitude surveys

Abstract

For decades, researchers have expressed concern that self-report racial attitude measures are vulnerable to distortion from pressures respondents feel to present themselves as unprejudiced. A common response to this problem is to measure social desirability separately from racial attitudes and control for its variance in statistical analyses. The present study is designed to test whether such controls are sufficient. Participants rated items from both racial attitude and social desirability scales in terms of the amount of pressure they would feel to respond in a particular way regardless of their true attitudes. Participants report significantly greater response pressure on racial attitude items than on social desirability items, and ratings on the two types of items have only moderately shared variance. Implications for controlling social desirability in racial attitude research are discussed.