Correcting for bias using multidimensional measures of Socially Desirable Response

We use multidimensional measures of socially desirable response (SDR) to reduce the bias introduced by respondents’ need for approval from interviewer (Impression Management — IM) or to perceive oneself favorably (Self-Deceptive Enhancement — SDE). Using data from a field experiment, we show that (1) SDE explains misreporting of compliance status, (2) controlling for SDE effectively eliminates the bias resulting from using wrongly reported compliance and (3) controlling for IM does not correct for the bias introduced by SDR.