3. Due to the small number of items and the heterogeneous content of different socially
desirable behaviors in different contexts low Cronbach’s alphas are not surprising (cf.
Cortina, 1993; John & Soto, 2007; Schmitt, 1996).
4. Removing items due to high correlations with IM might reduce content validity of the
leadership self-rating as well as convergent validity (e.g., correlations with conscientiousness
or agreeableness). The same should be true for SDE. Thus, items were not eliminated
(cf. Pauls & Stemmler, 2003).