| The effects of child sexual abuse (CSA) are well documented with harms that impact across every domain of a child's functioning, yet accurate data on the extent of the problem is lacking. This data is critical to inform policies and programs aimed at effective intervention. A fundamental first step to achieving this end is the development of a theoretically-based, valid and reliable measure of child maltreatment. However, few theoretically-based instruments are available and fewer still have been rigorously evaluated. The Childhood Experiences of Violence Questionnaire (CEVQ) was developed within an ecological-developmental framework as a population-based measure of youth victimization.;Preliminary evidence on the psychometric properties of the instrument suggests that it is a reliable and valid measure of child physical abuse (CPA) and CSA although the sub-types of CSA have not been evaluated. This thesis assessed the theoretical and psychometric properties of the CSA sub-types.;Methodological, data, and theoretical triangulation supports the developmental and ecological validity of the CSA measure, suggesting that youths are able to understand and respond appropriately to the CSA sub-types.;Psychometric evaluation supports the stability of CSA and severe CSA, with individual CSA sub-types and contingency questions exhibiting a moderate degree of stability. The face, content, construct, and criterion validity of the CSA sub-types demonstrate validity. However, greater variability was noted in the validity measures of each of the individual sub-types.;Significantly elevated scores for some behavioural and emotional disorders were noted for CSA, severe CSA and a severity score, suggesting a moderate support for a dose-response relationship.;The criterion validity and the sensitivity and specificity of CSA and severe CSA were better than the individual CSA sub-types when compared with clinician report suggesting that the CSA is a more effective measure than the individual sub-types.;The present finding suggests that the CSA measure of the CEVQ should be used as an epidemiological tool rather than as a diagnostic measure and the CSA sub-types are less reliable and stable than the composite measure. |
