Critical issues in construct validation: Comment on Boyle and Lennon (1994)

In a recent article in this journal, Boyle and Lennon (1994) examined the reliability and validity of the Personality Assessment Inventory (PAI). However, several methodological and conceptual limits of their study suggest numerous alternative interpretations of their data. Methodological problems include a confounding of age and clinical status in their study, their inclusion of invalid PAI profiles, and use of an unusual sample for their purposes. Conceptual issues raised address the nature of a moderator variable, the interpretation of coefficient alpha, the stability of psychopathology, and the application of factor analysis to address construct validation issues.