What it’s about

In this paper, we examine how a single creativity score in PISA can mask two separable dimensions of judged responses, and we propose generalized item response tree modeling to estimate appropriateness and originality components explicitly.

Abstract

Creative-thinking judgments in large-scale assessments often combine multiple evaluative criteria into one score, which can blur interpretation and measurement precision. This article focuses on PISA creative-thinking judgments and introduces a generalized item response tree approach to separate appropriateness and originality into distinct latent components. The model-based framework allows these components to be estimated jointly while preserving the structure of the observed scoring process. The paper illustrates how this decomposition clarifies what is being measured and supports more informative psychometric interpretation of creative-thinking performance.

How to access the paper

Go to the journal page for this article here.

Download the PDF here.