Three more attempts to prevent faking good in personality questionnaires

115 There is a good deal of evidence showing that personality questionnaires are fakeable. Impressive evidence was first given by Viswesvaran and Ones (1999), but also since then, in particular by Birkeland, Manson, Kisamore, Brannick, and Smith (2006), Deller, Ones, Viswesvaran, and Dilchert (2006), and Morgeson, Campion, Dipboye, Hollenbeck, Murphy, and Schmitt (2007a). In this paper, the phenomenon of faking good – especially in the context of job recruitment – is of prime interest, though faking bad also occurs within a clinical context (for a review, cf. Franke, 2002). Most studies on this topic use some kind of instruction to fake, which might be the reason why practitioners do not take these results into account. They argue that experimental behavior is different from behavior within a job recruiting procedure. Sometimes it is even argued that faking indicates a kind of social competence and is beneficial within recruiting (cf. for instance Marcus, 2003), or that faking is not harmful because everybody fakes and thus only the mean score is altered (cf. Kanning & Holling, 2001). Of course, there is no reason to trust so-called lying scales, given that these have proven to be fakeable as well (cf. Moorman & Podsakoff, 1992). Although it is easy to counter these claims with various arguments, their discussion is of little relevance to this paper and only two such arguments will be mentioned. Firstly, social competence in no way guarantees that the original personality traits, which the questionnaire aimed to measure, are still given satisfactorily, while loyalty to the company is undermined. Secondly, the differing behavior of different people is disregarded and therefore candidates who do not fake or who fake less are at a disadvantage. The question is no longer whether examinees actually fake in answering a questionnaire, but rather what psychological means are at psychology’s disposal to conceptualize personality questionnaires that are – for the most part – able to prevent an examinee from faking. In answering this question, the results of two experiments will be presented.