Constructs and operational definitions in psychology. When assessment misrepresents the phenomenon: ‘Alexithymia’ as a case study

Psychological models rely heavily on hypothetical concepts, such as empathy, emotion, anxiety, personality, and intelligence, to describe, interpret, and explain human behavior. Such concepts do not denote directly observable phenomena but help us make sense of the available behavioural and neurophysiological evidence on our mental life. A number of experimental psychologists take the view that these concepts are best understood as constructs based on operational definitions, similar to those proposed by Bridgman in physics. This view is called operationalism (or operationism) and it is typically considered a failed philosophical project. Yet, this is still prevalent in the practice of psychological research because it offers increased conceptual accuracy and reliability. A wellknown flaw of operational definitions is their intrinsic dependence on the way we measure the concepts we aim to define. If we use multiple measures, we end up with a harmful proliferation of constructs,which, in turn, can give rise to unwanted epistemological consequences.While recent research on the use of operationalism in psychology argues that it does not fall victim to these objections, in this paper, we rely on a specific example – the construct of alexithymia (i.e., a subjective condition characterized by the incapacity of some people to fully experience their own emotions) – to show that at least some operational definitions widely used in psychological research remain highly problematic and could lead to misrepresentations of phenomena. The discussion of this example will be followed by a general analysis aimed at determining which kinds of operational definitions in psychology are particularly at risk and why.