First we assessed bivariate associations between Black and White victims on timing, type, severity, and chronicity of maltreatment using t-tests for continuous variables and chi-square tests for categorical variables. Next, percentages and means of outcome variables between victims and non-victims, and Blacks and Whites were compared. Then, a series of sequential regression models (negative binomial, logistic, and ordinary least squares regression depending on the outcome, as described below) were tested to examine the main (Step 1) and interaction (Step 2) effects of childhood maltreatment and race on the outcomes.
