Abstract

The philosophy of measurement in the social and behavioral sciences is seen (from without) as typically following the representational viewpoint. However, in practice, this is not the case for the great majority of measures that are developed in this area. The paper surveys several approaches to measurement in the social sciences (i.e., Classical Test Theory, Guttman Scaling, Item Response Theory, Rasch Scaling, and Construct Modeling), as examples of measurement approaches in the area of psychometrics, and uses the foundational concept of a measuring system, as developed by Mari [1], to explicate the logic on which these approaches are based and thus enable a comparison with measurement approaches used by other fields such as engineering and physics. The paper uses the underlying concept of the standard reference set (one of the essential features of Mari's formalization) to show how the five approaches differ, and also how they are related. The importance of these differences, and the consequences for measurement using those approaches are also explicated and discussed. (C) 2013 Elsevier Ltd. All rights reserved.