Are improvements in assessment center construct-related validity paralleled by improvements in criterion-related validity?

Previous studies have found that factors that improved assessment center (AC) construct-related validity also had beneficial effects on criterion-related validity. However, some factors might have diverging effects on construct- and criterion-related validity. Accordingly, we followed recent calls to evaluate construct- and criterion-related validity of ACs simultaneously by examining the effects of exercise similarity on both aspects of validity within a single study. Data were collected in an AC (N = 92) that consisted of two different types of exercises. Convergent validity was better for similar exercises than it was for dissimilar exercises. However, regarding criterion-related validity, we did not find differences between similar and dissimilar exercises. Hence, this study revealed that improvements in AC construct-related validity are not necessarily paralleled by improvements in criterion-related validity