Social Consequences of Educational Measures: A Commentary

This commentary discusses Messick's (1995) unified validity framework, which broadly characterizes the interpretation of test scores in terms of their social consequences. Scores and their interpretation can lead to unintended consequences when they capture only part of the specified construct or characteristics unrelated to the construct. The evaluation of construct underrepresentation and irrelevance requires careful investigation and rational argument about the construct and its theoretical basis, as well as any planned uses, contexts, scores, or samples. Developers also validate an assessment for specific purposes, and users share responsibility for validation for any novel use or interpretation of scores. This commentary also considers the consequences of decisions based on assessments and the consequences of local and national norms. Impact and Implications The validity of educational measures depends on their score interpretation. If a measure's scores capture only part of the intended construct or include attributes unrelated to the intended construct, the use of the measure can lead to unintended consequences. Measure development and administration require a careful investigation of the intended construct and its theoretical basis, as well as a clear argument for its scoring and use in planned contexts and samples.