Examining the Reliability, Factor Structure and Criterion Validity of Marlowe-Crowne Social Desirability Scale

In self-report questionnaires, possess social desirability bias can be seen. Then, assessing and controlling social desirability is one of the methods to promote the findings based from such questionnaires. One of the most popular instruments that assesses this bias is 13-items Marlowe-Crowne Social Desir-ability Scale (MCSDS; Reynolds, 1982). How-ever, the Persian version of it does not have satisfactory reliability. In this study, using data from some previous studies among university students (N=1439) and data from a longitudinal study among them (N=90), the scale was re-validated. Explanatory and confirmatory factor analysis, correlation bet-ween items and total score, discrimination of the items, relaiability by internal consistency method and examinaing the criterion validity by compairing annony-mouse and named questionaires, all were done among the first testing sample. In addition, test-retest reliability testing was done on the longitudinal study. Explanatory factor analysis has shown that the scale consists of three factors. Confirmatory factor analysis confirmed the three-factor model among the entire sample. Howevere, focusing on the meaning factor leaded ambiguity for naming of the items at any factor leaded ambiguity for naming the factors. In addition, criterion validity, internal consis-tency, and stability of the factors, also discrimi-nation of the items, all questioned the reliability and the validity of the first and the third factors. It seems that they assess another things than social desirability. By omitting the first and the third factors, seven items were remained that are suitable to assess social desirability bias for the studies that use self-report questionairs.