Social, mental and physical function are major components of health outcomes and health related life quality, but the accuracy of function measurement is difficult to study rigorously. Internal Consistency Analysis (ICA) uses multiple sources of evidence from a survey interview to study the accuracy of a classification. It was developed to study function classifications for a general health outcome measure, the Quality of Well-being (QWB) scale. ICA is described and evidence of its utility in improving the classifications needed for the QWB is presented.