Development and psychometric evaluation of a teacher version of the Ontario child health study emotional behavioural scales (OCHS-EBS-T) for measuring selected DSM-5 disorders in elementary school-aged children

This paper describes the development and psychometric properties of a teacher version of the Ontario Child Health Study Emotional Behavioural Scales (OCHS-EBS-T) for dimensional measurement of six psychiatric disorders in elementary school-aged children based on DSM-5 criteria. Psychometric evaluation of the item and scale properties was conducted in a large, general population study of elementary teacher assessments of 3,072 children aged 4 to 13 years in 2,354 families in Ontario, Canada. Content validity, internal construct validity and internal consistency reliability of the six disorder subscales were assessed. Structural equation modelling was used to assess measurement invariance, internal and external convergent and discriminant validity. Confirmatory factor analyses confirmed unidimensionality of subscales and adequate item-fit to all scales. Measurement equivalence was demonstrated fully for four subscales and partially for two. Internal consistency reliability for all subscales exceeded 0.78, except for the conduct disorder scale in 12- to 13-year-olds. Evidence of internal convergent validity was demonstrated in all cases. Discriminant validity was demonstrated in 27 out of 30 correlation comparisons. External convergent and discriminant validity was demonstrated when comparing the OCHS-EBS-T to a parent/caregiver measure of disorders in 48 out of 60 correlation comparisons. All subscales independently predicted child mental health-related service contact.