Improving Global Measurement of Intimate Partner Violence

PROJECT SUMMARY/ABSTRACT Significance. Intimate partner violence (IPV), the most common form of violence worldwide, is associated with a wide range of negative health outcomes for victims and exposed children. Precise, valid, and transportable measures of IPV are needed to accurately capture the prevalence of IPV, assess the effectiveness of interventions to prevent and reduce IPV, and appropriately allocate resources to combat this pervasive global health issue. This project will address five limitations of current IPV measurement: measurement error, underreporting, lack of consensus in IPV domains and outcome construction, lack of equivalence across countries, and small item sets. Aims. This study aims to: (1) assess the performance of a quantitative method to control for measurement error and underreporting of IPV; (2) identify the best-fitting measurement model of IPV incorporating physical, sexual, and psychological subdomains in 45 lower- and middle-income countries (LMICs) and assess its comparability across countries; and (3) test the appropriateness of the model for higher-income countries (HICs) and assess the item functioning of IPV scales used in HICs to identify possible items to add to IPV scales currently used in LMICs. Data. This study will use data from the Demographic and Health Surveys, Multiple Indicator Cluster Surveys, National Intimate Partner and Sexual Violence Surveys (NISVS), and the European Union Agency for Fundamental Rights’ Violence Against Women Survey (FRA-VAW). Methods. For Aim 1, I will apply multiple overimputation to correct measurement error and underreporting in multipurpose surveys. For Aim 2, I will use confirmatory factor analysis (CFA) to test the fit of four models of IPV in LMICs and use multi-group CFA to assess their equivalence across countries. For Aim 3, I will test the fit of the model identified in Aim 2 in HICs and use item response theory to assess item functioning of scales used in NISVS and FRA-VAW. Impact. These investigations will inform recommendations to improve the precision, validity, and transportability of IPV measures, contributing to better surveillance and evaluation of prevention efforts. Training. This project will strengthen my subject matter knowledge of IPV measurement; develop key quantitative skills in overimputation, Bayesian modeling, item response theory methods, and structural equation modeling techniques; and hone my ability to communicate scientific findings. Formal training, dissemination activities, and the mentorship of Drs. Cari Clark (sponsor), Regine Haardoerfer (co-sponsor), and Robin Richardson (mentor) will enable the next steps in my career as an independent researcher in the field of violence prevention.