Three regression approaches are compared and the authors review the Literature on return rates. Recent Research on Muiled Questionnaire Response Rates Robert M. Baumga rtner Thomas A. Heberlein zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA In 1978, we published an article (Heberlein and Baurngartner, 1978) that was a quantitative review of 214 mailed surveys. Since that time, about forty addi- tional studies of mailed surveys, including two replications of our original work, have been conducted. This chapter reviews these new studies to access advances in research on mailed questionnaires over the last five years. In our original study, we coded seventy-one factors reported in method- ological studies of mailed questionnaires and examined their effect on response rates. Through analysis, we reduced the seventy-one factors to ten variables, which predicted 66 percent of the variability in final response rate, Subse- quently, Goyder (1982) replicated our analysis and procedures on an intended data set that included more than 330 surveys, and Eichner and Habermehl (1981) attempted to replicate our model on an independent sample of 152 mailed surveys from Austria and West Germany. We will use the ten variables in Table 1 to organize the review. Research that does not fall into these categories will be discussed separately. Sponsorship In our original analysis (Heberlein and Baumgartner, 1978), studies sponsored by a market research organization had a 10 percent lower response U. C. Lockhart (Ed.). Making EJeche Use zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA of Mailed Questionnaires. New Directions for Program Evaluation, no. 21. San Francisco:Jossey-Bass, March 1984 65