Factor analysis as a tool in primary care research

Research in primary health care is increasingly relying on questionnaires and surveys to address relevant research questions. In many cases, a variety of items are assessed representing attitude sets or multidimensional constructs. Factor analysis is an effective and efficient method for identifying underlying dimensions in a group of variables. In addition, factor analysis allows the investigator to reduce a large set of original variables to a smaller set of new composite factors for use in future investigations, with a minimum loss of information. The issues and methods of factor analysis are described and illustrated with data from an investigation of a medical school's attitudes toward gifts from industry.