Psychological Constructs as Organizing Principles

Abstract

Klaas Sijtsma has suggested that psychological constructs, such as those invoked in the study of intelligence, personality, and psychopathology, should be understood as organizing principles with respect to elements of behavior, including item response behavior. In a discussion in the journal Psychometrika, Sijtsma (Psychometrika, 71(3), 451–455 (2006)) contrasted this position with the common cause interpretation of Item Response Theory (IRT) models and the associated theory of validity that I had articulated some years earlier (Borsboom, Psychological Review, 111(4), 1061–1071 (2004)), arguing that this theory of validity was far too strong given the immature status of psychological constructs. In the present chapter, I present an alternative understanding of IRT models in terms of psychometric networks, which is inspired by Sijtsma’s idea of constructs as organizing principles. From the weak premise that psychological constructs organize behaviors, in the sense of identifying behavioral elements that structurally hang together, in the present chapter, I show how one can build up a psychometric approach that can motivate and guide the use of tests in psychology in the absence of strong common cause interpretations.