The challenge of defining and interpreting dimensionality in educational and psychological assessments

In psychology and education, attributes often lack a clear or universally agreed-upon dimensional structure. Constructs, such as intelligence, personality, teacher quality, or school climate, are typically assessed under seemingly unified labels. However, these labels often conceal complex bundles of concepts, ranging from simple factor lists to elaborate theoretical models outlining their relationships. Consequently, the assumption that multiple dimensions are required to capture this intricate complexity raises questions about how many dimensions to consider and how to interpret them. Previous literature on dimensionality has predominantly focused on the coordination between theoretical and statistical perspectives. On the one hand, researchers explore the theoretical attributes hypothesized to underlie assessment results. On the other hand, they examine the dimensional structure as a fundamental assumption underlying the statistical models used to generate results. However, assessments are often subject to constraints driven by time, resources, legal or regulatory requirements, and the users' capacity to comprehend the results. These constraints operate independently of theoretical specifications or the ideal fit of statistical models. This paper argues that an additional perspective of use is crucial for addressing the challenges posed by dimensionality. Integrating these perspectives-theory, statistics, and use-is essential to understand comprehensively the complexity surrounding dimensionality. These perspectives represent a descriptive account that considers the constraints imposed by different groups involved in designing an assessment instrument, including substantive researchers, psychometricians, and users. Coherence among these strands is essential, aligning the guiding theory for assessment design, the statistical models used to produce results, and the practical utilization of those outcomes.