摘要：Different means can be applied to assess noncognitive personality aspects: projective, semi-projective, self-report, and objective. However, so far little attention has been paid towards the different fakability of these methods. The present study investigated this question with different achievement motivation instruments. The instruments were randomly administered to three student groups: fake bad (n = 41), fake good (n = 37), and control group (n = 41). The faking groups were given specific faking instructions while the control group only received the standard instructions. All instruments were applied computer-assisted. The results show that all tests are fakeable with the exception of the objective measure which could not be faked good as was expected. The effect sizes (d) ranged from .10 to 2.36. Cut-off scores for the detection of faking were computed based on sensitivity as well as specificity. Moreover, they were tested within a second student sample (n = 123). Sensitivity and specificity values are reported. The practical implications for test authors and practitioners are discussed.
