Relevance & Research Question: The Marlowe-Crowne Social Desirability Scale (MC-SDS, Crowne & Marlowe, 1960) is the most widely used tool for measuring social desirability (SD). It was originally developed as a paper-and-pencil questionnaire using a student sample. Since its publication in 1960, most studies of the MC-SDS and its short forms have used paper-and-pencil versions and relied on student samples. The only study (known to the authors) specifically aimed at testing the psychometric properties of the scale in internet format was conducted on an Icelandic snowball sample. The analysis produced compatible results to prior paper-and-pencil studies, in spite of differences in language, mode and sample (Vésteinsdóttir, Reips, Joinson & Þórsdóttir, 2012). Although this can be taken as evidence that the psychometric properties of the scale are fairly stable across modes, this evidence is limited to an Icelandic nonprobability sample. To assess the generalizability of the results obtained on the Icelandic sample, the current research evaluates the psychometric properties of a MC-SDS short form in a Dutch internet sample.
Methods & Data: Use was made of data from all five waves of the Personality study of the LISS panel, administered by CentERdata, which contains a 10 item version of the MC-SDS. The LISS panel is based on a probability sample and is therefore a more representative internet sample than has been used before.
Results: Confirmatory factor analysis (CFA) and reliability analysis were used to evaluate the psychometric properties of the MC-SDS short form. Reliability estimates were low but little higher for the five waves of the Dutch sample (.48-.54) than for the Icelandic sample (.43). Using CFA an adequate fit was obtained for the scale in all waves of the Dutch sample and the Icelandic sample. Moreover, measurement invariance was established between the Icelandic and the Dutch samples. Measurements obtained with this short form of MC-SDS are however not reliable enough and the validity of some of the items is questionable.
Added Value: Examining the psychometric properties of the MC-SDS in internet format is an important step in the process of finding suitable ways of dealing with SD in internet surveys.