Factor analysis as a tool in community diagnosis.

Abstract

For the family physician, community diagnosis can be likened to individual diagnosis by the application of methods of information acquisition and analysis. This paper describes the technique of factor analysis, applied to data from an urban inner-city area, for reducing a multitude of factors to a manageable number without an appreciable loss of information. The variables were selected from four broad categories which have been shown to influence both health states and health care needs: demography, socioeconomics, social disorganization and morbidity/mortality. A correlation matrix based on 40 of the most accessible indices served as input for a principal axes factor analysis. Six factors accounted for 74% of the total variance and were interpreted as: 'Poverty and social disorganization', 'Distribution and problems of the elderly', 'Ethnic composition', 'Fertility', 'Infant mortality' and 'Foetal mortality'. Representative indices from each factor were then mapped to identify and display census tract differences. Finally, the application of the factors to a specific community illustrated important differences within the community and identified areas of high risk and need.