To discover factors which contribute to a high response rate for questionnaire surveys, the preferences of 150 college teachers and teaching assistants were studied. Four different questionnaire formats using 34 common items were sent to the subjects: open-ended; Likert-type (five points, from "strong influence to return," to "strong influence not to return"); checklist; and rank ordering of five to eight factors. Subjects were also asked how many questionnaires they had received, and returned, in the past year. When instructed to choose one of the four questionnaire formats and complete it, 54 subjects (36%) completed and returned a questionnaire; 22 chose the checklist, 17 the Likert-type, 14 the open-ended, and 1 the ranking. Characteristics of each type of questionnaire which influence rate of return--such as physical characteristics, content, and administration methods--were analyzed from the answers received. Subjects preferred clarity and brevity; familiar but not overly personel subject matter; and convenient methods of completion and return. A humorous followup reminder, if any, was preferred. Respondents also wished to know the purpose of the research and the researchers, and to be assured of confidentiality. Non-response was predicted for questionnaires which did not meet these criteria. (GDC)