Racial/ethnic Differences in Expanded ACEs: Differential Item Analysis Using the Rasch Model

Background: Adverse childhood experiences (ACEs) are associated with poor health outcomes and risky health-related behaviors. There are numerous measures of adverse childhood experiences and findings to date have varied across studies. No item-by-item analysis has been conducted to determine how likely it is for individuals from distinct groups, who possess similar abilities, to respond differently to items on the Philadelphia (PHL) ACEs Survey created by Cronholm et al. (2015). Objectives: To determine whether there is measurement invariance in conventional and community-level ACEs items on the PHL ACEs Survey in non-Hispanic Black, non-Hispanic White, and Hispanic adults. Methods: Participants aged 18-34 were recruited via ResearchMatch. Participants completed demographic items and the Philadelphia ACEs Survey on REDCap. A differential item functioning (DIF) analysis was completed using a Rasch model, DIF contrast values were identified, and a Mantel-Haenszel test was completed with a Benjamini and Hochberg adjustment to identify DIF. Results: The sample (N=144) was roughly one third Black, one third Hispanic and one third White adults. Two conventional ACE items exhibited differential item functioning: 1) Did you live with anyone who was suicidal? was easier for Hispanic adults to endorse than Black adults; 2) Did you live with anyone who served time or was sentenced to serve time in a prison, jail, or other correctional facility? was easier for Black adults to endorse than Hispanic adults. One community ACE item: How often did you feel that you were treated badly or unfairly because of your race or ethnicity? was easier for Black adults to endorse than Hispanic and White adults. Conclusion: This is the first study in which a DIF analysis was conducted on the Philadelphia ACEs Survey to examine measurement invariance in Black, White and Hispanic adults. While most items functioned in the same way across groups, three items performed differently between groups. Due to the historical and ongoing systemic racism in the United States, the items that showed DIF (discrimination, suicidal household member and incarcerated household member) can most likely be classified as benign DIF (true differences) which does not impact measurement quality.