The Next Generation in the Measurement of Adolescent Relationship Abuse

Statement of problem. For more than a decade, NIJ has been committed to improving measurement of adolescent relationship abuse (ARA). There is a general consensus that current ARA measures do not adequately evaluate this phenomenon. One major shortcoming of the most common ARA surveys is that behaviors are measured without regard to context, and as a result, typical adolescent behaviors cannot be easily differentiated from abusive ARA behaviors. Moreover, existing surveys do not effectively characterize variations in the nature and severity of ARA. The primary objective of the proposed research study is to demonstrate that two contextual features, intention of the perpetrator and the impact on the victim, are key to improving the precision of ARA measurement, and future ARA surveys must incorporate these attributes.
Building on the NIJ Concept Mapping Study as well as recent survey innovations, the proposed research study will develop an ARA survey that can differentiate developmentally typical behaviors from ARA as well as identify different forms and severity of ARA across various dating contexts.
Subjects. 1230 adolescents and young adults aged 12-24 years will be recruited for the proposed study from three public middle schools, one public high school, and one public regional university in Northeastern United States. The sample will be 50% female, 45% White, and approximately 50% Black or Latino (school sample is 32% Black, 10% Latino; university sample is 16% Black, 32% Latino). An expert panel of 15 scholars and practitioners will give feedback on the preliminary survey.
Design and Methods. The methodology follows the National Institutes of Health Patient-Reported Outcome Measurement Information System (PROM IS) guidelines for instrument development. There are three objectives of the proposed project: 1) Item Development (Jan-Aug 2017) using focus groups (12 groups, n=10 per group); 2) Item Testing (Sep 2017-Jun 2018) using a Q-sort (n=150), expert panel review (n=15), and cognitive interview survey pretest (n=90); and 3) Pilot Test of ARA survey (Jun 2018-May 2019) with 900 participants to examine psychometric properties: reliability, factor structure, item discrimination, and validity.
Analysis. Qualitative analyses will be used to evaluate focus group, expert panel, and cognitive interview data. Q-sort data will be analyzed with factor analysis. Pilot test data will be analyzed with factor analysis, item response theory, and correlation analysis. Products/Reports. Quarterly, semi-annual and final reports will be prepared along with scholarly papers for publication. All data collected with be archived with the National Archive of Criminal Justice Data. ca/ncf