Health Disparities
Revealing Discrepancies: Exploring Racial Differential Item Functioning in Volunteering Activities and its Structural Underpinnings Yusuf Ransome* Ester Villalonga-Olives Abdolvahab Khademi Inez Adams Candace Hall
Introduction: In past quantitative studies, we identified racial Differential Item Functioning (DIF) in social capital items for Blacks and Whites, posing a challenge in examining racial disparities in social capital and its health-related implications. Few studies have qualitatively explored the roots of DIF in social capital questions. Our study aims to quantify DIF in questions assessing volunteering activities, a social capital indicator, and qualitatively investigate its underlying causes.
Methods: We conducted a mixed methods study to explore racial differences. We used longitudinal data (1995-2016) from the Midlife in the United States study (n= 6695). Volunteering activities, measured by monthly hours spent volunteering for organizations, were analyzed for DIF through Item Response Theory (IRT) and structural equation modeling. Additionally, we delved into the root causes of DIF through qualitative interviews (N=45).
Results: DIF was observed in IRT results for school volunteering items, with race affecting the conditional probability of volunteering. Black individuals consistently showed higher likelihood of choosing specific responses for volunteering in social organizations (P(χ2,2) = 0.00). Longitudinal DIF revealed a lack of metric and scalar invariance over time for White individuals. Qualitative findings suggested similar interpretations of volunteering questions and activities among Blacks and Whites, highlighting location distinctions, especially in urban versus suburban settings.
Conclusion: While the quantitative results indicated the presence of racial DIF, the qualitative study offered more insight into the underlying factors contributing to DIF. It proposed that the quantitative findings might be elucidated by the observation that Blacks are more commonly situated in urban settings, whereas Whites tend to be in suburban settings. Our study reveals the value of integrating qualitative insights to understand the nuanced factors contributing to DIF.