The measurement of opinions helps to indicate students', teachers', and parents' satisfaction with school programs, as well as students' self concept and affective behavior. In choosing a sample, several issues should be considered: sampling error, sample size, selection procedure, and acceptable response rate. Follow-up survey procedures and questionnaire validity are also important concerns. Guidelines for item construction include: (1) word questions so as to elicit the same interpretation from every respondent; (2) use simple, familiar language; (3) avoid offensive questions; (4) avoid questions which suggest the socially desirable response; (5) include only one issue per item; and (6) do not confuse factual or cognitive items with opinion or affective items. Rating scale item formats are useful in measuring attitudes, opinions, and feelings; a variety of response alternatives may be appropriate. Pilot testing is recommended to insure the validity of the questionnaire. The decision of whether to use hand-tabulated or computer-processed analysis of data should be made before the questionnaires are completed. (GDC)