The psychometric properties of childhood physical and sexual abuse measures in two Canadian samples of youth and emerging adults

TL;DR

This study evaluates how well different questionnaires measure childhood physical and sexual abuse in Canadian youth and young adults.

Contribution

The study provides new evidence on the reliability and validity of the CEVQ-SF and CCHS-CSA measures compared to the CTQ in two Canadian samples.

Findings

Prevalence of child physical abuse ranged from 12.5% to 41.4%, and child sexual abuse from 5.8% to 34.3% across two samples.

The CEVQ-SF showed good internal consistency for physical abuse and acceptable consistency for sexual abuse in the samples.

Agreement between CEVQ-SF and CTQ was moderate-to-fair for physical abuse and substantial for sexual abuse in one sample.

Abstract

Child maltreatment is prevalent in Canada; how we measure it varies. The objective of the current study was to examine the psychometric properties of the Childhood Experiences of Violence Questionnaire Short Form (CEVQ-SF) physical and sexual abuse measures and of the Canadian Community Health Survey (CCHS) 2-item sexual abuse measure, compared with the Childhood Trauma Questionnaire (CTQ) in two samples of adolescents and young adults.

Retrospective, self-reported child abuse history was collected in the British Columbia Healthy Connections Project (BCHCP) and in the Well-Being and Experiences (WE) Study. Internal consistency, criterion validity, and construct validity were examined.

Across both samples, the prevalence of child physical abuse (CPA) and child sexual abuse (CSA) ranged from 12.5% to 41.4% and from 5.8% to 34.3%, respectively. Internal consistencies were good-to-acceptable for CPA using the CEVQ-SF in the BCHCP (α = 0.83) and the WE Study (α = 0.79) and for CSA using the CEVQ-SF in the WE Study (α = 0.68). For CPA, in both studies, the highest agreement—moderate-to-fair—was between CEVQ-SF severe CPA and CTQ moderate CPA: κ=0.63 (BCHCP) and κ= 0.35 (WE Study). For CSA, agreement with CTQ moderate cut-offs was substantial in the BCHCP (κ=0.77) and fair in the WE Study (κ=0.37).

Our findings support current and future use of the CEVQ-SF for CPA, and for CSA, using both the CEVQ-SF and the CCHS-CSA measure, given that they had good psychometric properties when administered to two samples of adolescents and young adults.

Figures