The Impact of Item Parceling Ratios and Strategies on the Internal Structure of Assessment Center Ratings

Abstract

Abstract. The aim of the present study was to investigate whether using item parcels instead of single indicators would increase support for the factorial validity of assessment center (AC) ratings in factor analytic applications. Factor analytic analyses of AC ratings are often plagued by poor model fit as well as admissibility and termination problems. In the present study, three purposive item parceling strategies, in conjunction with three parceling approaches (specifying different ratios of indicators to dimensions), were investigated in relation to five confirmatory factor analysis specifications of AC ratings across two AC samples (Sample 1: N = 244; Sample 2: N = 320). The findings were equivocal across the two samples. Nonetheless, a three-parcel approach using a factorial allocation strategy performed better that a one-parcel approach (akin to the postexercise dimension rating).