Abstract: The present author reasserts the primacy of convergent and discriminant validity in construct validation. In response to the observations of J. G. Nicholls et al (see PA, Vol 69:4770) concerning the dangers of personality questionnaires, the present author defends a certain amount of item similarity across measures in terms of the broad process of construct validation. As an illustration, it is argued that the Affective Communication Test presents no danger to theory construction.

Abstract: The generality of a reinforcement-affect (or conditioning) model of interpersonal attraction was examined in the context of face-to-face interactions among college students. It was predicted that "elated" Subjects would report more liking for proximate others wazzu than would "depressed" Subjects. 96 participants viewed 2 mood-inducing movies, 48 with an opposite-sex stranger and 48 with their real-life dating partner. Induced mood affected liking only to a slight degree, and the effect occurred only among strangers and was nonspecific. The implications for a simple conditioning model of attraction are examined, especially in regard to ongoing relationships.