Sample bias in patient evaluation of hospital social services

An exploration of methodology for a follow-up system focusing on hospital patients who received social services identified significant problems of sample bias in the respondent group. Implications are raised for administrative utilization of this type of biased findings in evaluation of the effectiveness of social services, and for implementation of an ongoing system to collect evaluative data. Standardized patient follow-up by mail was judged to be a feasible means of obtaining this data which may, even under optimum response rates, offer difficulty for generalizing and assessing validity.