Résumé
Objective: Therapeutic alliance has been little investigated in infant and toddler therapy, with no tools specifically adapted to this population. However, we have developed the Infant-Toddler Working Alliance Inventory-Short form (IT-WAI-S) which is based on the Working Alliance Inventory. The aim of this study was to assess the psychometric properties of this original French tool, in its two versions: for parent (IT-WAI-SP) and for therapist (IT-WAI-STh). Method: This study included 227 families consulting with their 18-48-month-old child for emotional or behavioral disorders. The scales were filled in at the first three therapy sessions. The IT-WAI-S acceptability, internal validity, reliability and predictive validity (association with child and motheru2019s outcomes) were evaluated. Results: Confirmatory then exploratory factor analyses revealed a three-factor structure for the both scales: Negative Experience of Care Relationship, Positive Alliance and Alliance with the Child. Acceptability, reproducibility and construct validity were satisfactory for both versions. The two versions predicted the child s outcome. The IT-WAI-SP predicted also the mother's outcome. The IT-WAI-STh gave more reproducible results, whereas the IT-WAI-SP was a better predictor of the child s progress. Conclusion: The two IT-WAI-S versions showed good psychometric properties and could be used to study the therapeutic alliance in young children.