Instruments to collect data about families are often administered to all or some individuals within the family. Researchers may wish to use these individual scores to describe the family. The purpose of this article is to describe the special issues with aggregation of data when only a small number of family members are used as respondents. A refinement of the definition of aggregation for family researchers is proposed to assist researchers to focus on specific issues when data are to reflect subgroups within the family. A few examples of changes in findings are reported to illustrate the effects of different aggregation schemes when two members of the family are used as respondents.