Measuring Within- and Between-Group Agreement: Identifying the Proportion of Shared and Unique Beliefs Across Samples

Abstract

Questions about culture sharing and uniqueness can be addressed by measuring the amount of agreement within and between groups. This article describes methods to measure withinand between-sample agreement with particular emphasis on finding, using, and interpreting agreement. The discussion focuses on the case of a single set of related questions asked of two or more groups of informants, but the methods apply to the more general case with a set of related variables recorded for two or more sample units (e.g., culture traits or inventory materials). Methods are illustrated with Weller and Baer's study of withinand between-group agreement for four Latino samples for beliefs about each of five illnesses (AIDS, diabetes, the common cold, empacho, and mal de ojo ).