Research and instructional role strains are identified, based on the use of large-scale mail surveys for college self-study that are used to teach introductory research methodology. The two organizational hierarchies, the research model and the instructional model, are examined. Experiences at Northwestern University, Northeastern Illinois University, and the University of Chicago illustrate how relationships are manipulated in response to different situational constraints. Critical resources identified for adequately gathering data are time, funding, and number and skills of students enrolled. The instructor acting as principal investigator must relinquish some control over the research process to both the university administration and student/staff. Accountability standards must be maintained or deviations explicitly noted concerning sampling, pretest, design of questionnaire, questionnaire distribution and return followups, and quality control of data processing. Measurement of research quality and instructional quality are examined. Recommendations are offered to reduce role strains and improve research and instructional relationships in subsequent experiential approaches to teaching survey research. (Author/SW)