Attrition biases in a study of Euro-American and African-American marriages

Based on survey data from 174 Euro-American and 199 African-American newlywed couples, this study analyzed attrition biases by comparing first-year responses of couples who stayed in the study into its third year (133 Euro-American and 115 African-American couples) with responses from the initial sample. Stayers--who were more likely than leavers to be better educated, wealthier, and Euro-American--tended to report happier, more affirming, more communicative marriages. For stayers, compared to a random subsample of the original sample, first-year marital happiness also correlated significantly less strongly with first-year reports of receiving affirmation from a spouse, having an unsupportive spouse, and engaging in marital conflict. Further, race differences in predictors of happiness for the initial sample were not evident among stayers, perhaps due to smaller variances in reported marital happiness and frequency of conflict for African-American stayers compared to African Americans in the original sample. Methodological implications for cross-cultural longitudinal studies are discussed.