A Composite Index for Statistical Inference for Ranking Metropolitan Areas

ABSTRACT The construction of a composite index is described to rank U.S. metropolitan areas for educational attainment. The suggested methodology departs from traditional rank sum methods in that this approach utilizes data resulting in a continuous scale, whereas in the rank sum method, no use is made of the actual values of the data, but merely their rankings. The proposed index circumvents several practical problems because the numbers generated are metric measures allowing the use of arithmetic and statistical operations. Data used are pupil/teacher ratio, an effort index, and academic options in higher education. Comparisons are made across areas by population size and by geographic region.