Research based on scientific realism should not make preliminary assumptions about mathematical structure representing human behavior: Cronbach and Gleser’s measure as an example

Abstract

Phenomena in the world we study can often be described by various mathematical structures. If a psychologist who studies human behavior takes a realist position, they should not choose a mathematical structure that represents this behavior without examination as to whether the phenomenon could be represented by this structure, but they should eventually choose the mathematical structure after thorough reasoning based on full knowledge of the problem. Using Cronbach and Gleser’s measure for assessing the similarities between psychological profiles, I show that psychologists often simply copy the mathematical structure used in other research without thoroughly reasoning about the problem. As Arocha (2021) shows, researchers should prefer approaches that include no unjustified assumptions about the mathematical structure that represents the behavior.