Social Desirability Amongst Chinese Teachers.

Abstract

Research has suggested that self-reported responses on surveys can be affected by a participant's tendency toward social desirability, which would prevent them from revealing their true feelings or behaviors. Researchers should provide evidence that their results have not been affected by socially desirable responses using the Marlowe-Crowne Social Desirability Scale (MC-SDS). Past research has used the 33-item original form and 13-item short form of the MC-SDS, although a few researchers have found questionable validation of the 13 item MC-SDS in several populations. Traditional factor analysis failed to converge on a factor structure. Therefore, the current research was conducted using a Rasch dichotomous model analysis on the original 33-item MC-SDS, a 20-item MC-SDS, and the 13-item MC-SDS. Findings indicated that the 33-item MC-SDS had several overlapping items, the 20-item MC-SDS provided a somewhat meaningful measure of social desirability, and the 13-item MC-SDS did not provide sufficient item distribution or separation to produce a meaningful measure. A researcher should check on the factor structure of the MC-SDS when using it in their research, especially with different cultural populations.