Is Psychometrics So Useful for Academic Psychology?

Abstract

Psychological theories regarding ability and personality traits often rely on the results of psychometric modelling. The latter is assumed to link responses to test items to an unobserved 'construct' (trait, ability), which is 'modelled' from the test data. However, does the agreement between the data and the model indicate that the model represents a psychological construct? To what extent is ‘psychometric modelling’ modelling in the general scientific sense of the term? The validity of using modelling data to understand psychological phenomena depends on the answer to these questions. The article analyses the logic of psychometric modelling in comparison with modelling in other sciences and argues that psychological phenomena as a subject of modelling are not involved neither in the construction nor in the correction of models. The problem of unjustified interpretations of modelling results in psychology and their undesirable consequences for psychological theory is raised. At the same time, the use of psychometric modelling for human resource decision-making is still waiting for its evaluation.