Sncio-Eron. Plan. Sri. Vol. 6, pp. 421-429 (1972). Pcrgamon Press. Printed in Great Britain zyxwvutsrqponmlkjihgfedcbaZYXW MEASUREMENT TESTS FOR EVALUATION OF SOCIAL INDICATORS* EMIL E. MALIZIA Department of City and Regional Planning, University of North Carolina, Chapel Hill, North Carolina zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO (Received zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ 18 October 1971) An evaluation methodology is developed in this paper based upon two criteria: corres- pondence and correctness. Four qualitative tests-construct-indicator correlation, construct dimensions-referents correlation, comparability of units and levels of analysis, and utilization of explicit weighting variable-apply the correspondence criterion. The correctness criterion is used in one test in order to compare relative errors in indicator and referents with a quantitative analysis of cumulative error. 1. INTRODUCTION THE INTEREST in social indicators has been increasing over the past decade paralleling the concern for analysis and evaluation of public policy. t While most work has been conceptual, the construction of systems of social indicators has recently begun.: The question of evaluating the quality of a particular social indicator or a system of indicators has been raised. However, the only suggested approach is to evaluate the potentia1 usefulness of social indicators to their consumers-public policy makers, managers and administrators in government and citizens. The purpose of this paper is to provide an alternative evaluation methodology by examining the information content of a social indicator or indicator system. The methodology represents an attempt to evaluate social indicators by analyzing the validity and reliability of the measurement process employed in indicator construction. A theoretical construct can be defined as an idea or abstraction which identifies some phenomenon in the real world. The level of material welfare in metropolitan areas, the amount of underemployment among Blacks, the degree of pluralism in political processes or the quality of the natural environment serve as examples. The theoretical constructs discussed in the social indicator Literature are ones which have policy relevance. They can be viewed as target or output variables. AlternativeIy, an empirical referent is the quanti- iative measure of zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA some phenomenon. Available data and statistics are empirical referents. A social indicator provides a linking function by utilizing empirica referents to measure some theoretical construct. Specifically, a social indicator can be defined as an empirical referent or, more typically, a combination of referents which quantitatively measures a * This research was supported by the Department of Housing and Urban Development, Contract No. H1216. Integrated Municinal Information System. Charlotte, North Carolina. t For bibliographical information on social indicators, see Beal [3] and Agocs [l J. 1 Work is proceeding in many government agencies, in universities and in research institutions, for example, The Urban Institute, Washington, DC. and National Bureau of Economic Research, Inc., New York.