Psychometric evaluation of the Ontario Child Health Study Emotional Behavioural Scales (OCHS-EBS) in children and youth from an outpatient mental health setting

This paper describes the psychometric properties of the Ontario Child Health Study Emotional Behavioural Scales (OCHS-EBS) for dimensional measurement of symptoms of seven common DSM-5-based psychiatric disorders in children and youth aged 4 to 17 years receiving mental health services from a large hospital-based outpatient service in Ontario, Canada. Psychometric evaluation used data collected about children and youth aged 4 to 17 years receiving outpatient mental health services. Caregivers reported on children and youth aged 4 to 17 (n = 3070) and youth aged 12 to 17 years provided self-reports (n = 1959). A sample of 99 youth and caregivers was used to estimate 7- to 14-day test-retest reliability. Internal consistency, test-retest reliability, structural, convergent and discriminant construct validity and measurement invariance by sex and age group were assessed using structural equation model (SEM)-based confirmatory factor analysis. External construct validity was assessed using independent sample t-tests and SEM analysis to evaluate known-groups validity. Internal consistency and test-retest reliability coefficients for all scales exceeded 0.75. Confirmatory factor analyses supported the hypothesized 7-factor structure. Internal convergent validity was demonstrated for all scales. Discriminant validity was supported in 105 out of 108 comparisons. Measurement invariance was demonstrated fully by sex and partially by age group. Known group differences were as hypothesized. The caregiver- and youth-reported OCHS-EBS provide reliable and valid dimensional measurement of seven DSM-5 disorders in clinical populations. Scores can be compared with norms from the general population for score interpretation and clinically-significant change thresholds facilitate the evaluation of change over time in clinical research and randomized control trials.