Units of Analysis, Measurement Theory, and Environmental Assessment

Abstract

The ecological fallacy is the error of interpreting variation among ecological entities as though it applies to individuals. It also is an error to interpret variation among individuals as though it applies to ecological entitles, and the theoretical and empirical ramifications of the latter error have been explored in a series of articles. A recent article by Levine commented extensively on this series and criticized some of its conclusions. The present article argues that Levine's article is incomplete in its application of measurement theory and that it overemphasizes an example that makes an inappropriate postulation about the distribution of universe scores. However, these disagreements are less important than areas of agreement. A substantial consensus exists about choice of units of analysis, usefulness of generalizability theory, and procedures for estimating the reliability of environmental measures.