Franz Pauls, Monika Daseking
The results provided structural support for the retention of the Auditory Working Memory index (AWM) as a structurally identifiable yet psychometrically redundant scale, whereas a subordinate factor for the Quantitative Reasoning index (QR) was completely unviable due to estimation failures and non-convergence across both model frameworks. A hierarchical model solution confirmed that the Nonverbal Index (NVI) represents a valid multidimensional construct, though model-based reliability analyses indicated that its underlying specific subscales lack the structural stability for independent interpretation beyond the NVI composite score. Furthermore, model-fit evaluations established that the General Ability Index (GAI) and the Cognitive Proficiency Index (CPI) are best represented by a second-order oblique four-factor model solution. While the GAI and CPI are reflected by plausible latent dimensions, their profound statistical proximity to the Full Scale Intelligence Quotient (FSIQ) and the underlying subscales' lack of sufficient statistical stability underscore the risk of profile misinterpretation.
INTRODUCTION: Aligning with the factor-analytical procedure already established for the original test version, the present study examined the structural validity and diagnostic viability of the German Wechsler Intelligence Scale for Children-Fifth Edition (WISC-V) ancillary index scales on an extended standardization sample of children and adolescents aged between 6 and 16 years (N = 1,466).
METHODS: Representing the first comprehensive investigation into the ancillary index structure of the German WISC-V, competing hierarchical and bifactor model solutions were systematically evaluated via confirmatory factor analyses (CFA) and supplemented by model-based reliability analyses.
RESULTS: The results provided structural support for the retention of the Auditory Working Memory index (AWM) as a structurally identifiable yet psychometrically redundant scale, whereas a subordinate factor for the Quantitative Reasoning index (QR) was completely unviable due to estimation failures and non-convergence across both model frameworks. A hierarchical model solution confirmed that the Nonverbal Index (NVI) represents a valid multidimensional construct, though model-based reliability analyses indicated that its underlying specific subscales lack the structural stability for independent interpretation beyond the NVI composite score. Furthermore, model-fit evaluations established that the General Ability Index (GAI) and the Cognitive Proficiency Index (CPI) are best represented by a second-order oblique four-factor model solution. While the GAI and CPI are reflected by plausible latent dimensions, their profound statistical proximity to the Full Scale Intelligence Quotient (FSIQ) and the underlying subscales' lack of sufficient statistical stability underscore the risk of profile misinterpretation.
DISCUSSION: Consequently, study limitations and boundaries for evidence-based assessment are discussed, emphasizing that subsequent diagnostic decision-making in clinical practice based on the ancillary index scales alone should be approached with caution absent further cross-cultural and clinical validation.