Who Participates in Bereavement Research? A Review and Empirical Study

Despite high attrition rates in bereavement research, the issue of selection biases on health measures has been neglected. This raises doubt about many of the generalizations concerning health characteristics and recovery patterns among bereaved people. If it is the least distressed individuals who participate in bereavement research, such research would underestimate the impact of bereavement on health. If, on the other hand, it is the most distressed who agree to participate (e.g., because they want to talk about their loss), the health consequences of bereavement would be exaggerated. The first part of this article reviews attrition rates in bereavement research and demonstrates both the prevalence of low acceptance rates and the lack of knowledge about the bias due to such self-selection. In the second part, an empirical study is reported which provides evidence on the two conflicting hypotheses described above. The findings of our study show that depression does indeed affect willingness to participate but operates differently for males and females: Whereas widowers who were less depressed agreed to participate in an interview, the opposite was the case for widows: Those who were more depressed did so. This sex difference is explained in terms of sex roles in coping styles and norms for exhibiting emotions. Empirical studies of bereavement reactions need to take such selection biases in sampling into account.