Abstract
Three short forms of the Marlowe-Crowne Social Desirability Scale were constructed from the results of principal components analysis ( N = 399). Those subscales were compared with short forms developed by previous researchers who used the same methodology. Examination of the subscales indicated that 13 of the scale's 33 items were isolated by at least two of the three reported studies. Those items were used to construct a composite subscale, which appeared to offer a useful alternative to the full scale. Further analysis of the subscale's contents, however, raised questions about the dimensionality of the Marlowe-Crowne scale. Caution was urged in the use and interpretation of both the full inventory and the short form until the meaning of scale scores can be clarified.