Article révisé par les pairs
Résumé : The observation of child behavior has to be made in ecologically valid contexts. Parent-child interaction was thought to be the most suitable context, since empirical evidence displayed strong associations with children’s behavioral outcomes, psychopathology, social relationships and well-being. Using clinical data from 137 caregiver-child dyads, the main goal of the current study was to test the psychometric properties of an adapted version of the Crowell Procedure among preschoolers. Despite the interest that the Crowell Procedure has aroused, its psychometric properties remain relatively under-documented. This research aimed to study (1) the association between parental and child behavior, (2) the discriminant properties of the Crowell Procedure between preschoolers with a clinical level of externalizing behavior and non-clinical children and (3) the correlation between the Crowell Procedure and a behavioral checklist. The results support the consistency of both tasks and scales, the discriminant properties, external validity and fidelity of the coding system. The Crowell Procedure can therefore be used as an observational paradigm to assess both child and parent behavior in clinical and research contexts. The discriminant analyses revealed that the procedure was effective at differentiating children displaying a clinical level of externalizing behavior from normally developing ones.