A New Perspective on the Multidimensionality of Divergent Thinking Tasks

Abstract

In the presented work, a shift of perspective with respect to the dimensionality of divergent thinking tasks is introduced moving from the question of multidimensionality across divergent thinking scores to the question of multidimensionality across the scale of divergent thinking scores. We apply IRTree models to test if the same latent trait can be assumed can be assumed across the whole scale in snapshot scoring of divergent thinking tests and holds for different task instructions and varying levels of fluency. This way, multidimensionality can be explored across scale points of a Likert-type rating scale and also multidimensionality due to differences in number of responses of ideational pools can be assessed. It was found that evidence for unidimensionality across scale points was stronger with be-creative instructions as compared to be-fluent instructions which suggests better psychometric quality of ratings when be-creative instructions are used. In addition, latent variables pertaining to low-fluency and high-fluency ideational pools shared around 50% of variance which suggests both strong overlap and evidence for differentiation. The presented approach allows to further examine the psychometric quality of subjective ratings and to examine new questions with respect to within-item multidimensionality in divergent thinking.

A Monte Carlo study of IRTree models’ ability to recover item parameters

“…Similarly, this study limited its scope to the commonly employed MPP model on a five-point scale, but there are other existing applications of the IRTree framework that vary in both overall model arrangement and the types of scales to which they are applied. Models for binary-coded data (i.e., testing data coded for accuracy; Debeer et al, 2017), as well as a range of polytomous scales (Jeon and De Boeck, 2016;Dibek, 2019;Ames, 2021;Spratto et al, 2021) have been developed, some of which are structurally distinct from the MPP decision hierarchy (e.g., Forthmann et al, 2019). Even the MPP model has been adapted to scales with varying response option lengths, altering the number steps in the decision hierarchy to accommodate for the number of response choices available (Plieninger and Meiser, 2014).…”