Generalizing Mail SurveyInducement Methods: Population Interactionswith Anonymity and Sponsorship

Population specific interactions with mail survey response inducement techniques have long been suspected as a source of variation in the reported effectiveness of mail survey inducement methods. This study investigated such interactions by use of a 2 x 2 x 110 randomized block experimental design involving 23,350 survey subjects. Highly significant, as well as substantial interactions were found between population factors and manipulations of survey sponsorship and anonymity assurance. Income level, education level, population flux, and regional sponsor loyalties were specifically linked to interactive effects with the experimental manipulations. Wesley H. Jones is Assistant Professor of Business Administration, University of Kentucky, Lexington, Kentucky. The author wishes to thank Dr. Joseph Guiltinan, University of Kentucky, and the Department of Parks, Commonwealth of Kentucky for their assistance and support. Public Opinion Quarterly ? 1979 by The Trustees of Columbia University Published by Elsevier North-Holland, Inc. 0033-362X/79/0043-0102/$1.75 This content downloaded from 157.55.39.55 on Sun, 12 Jun 2016 06:02:39 UTC All use subject to http://about.jstor.org/terms MAIL SURVEY INDUCEMENT METHODS 103 lations. University sponsorship has generally been found to be superior to government agency or private firm sponsorship. However, findings reported by Scott, by Sudman and Ferber (1974), and Houston and Nevin (1977) suggest that sponsorship may also interact with population characteristics. Concern with the population interaction issue and the related lack of generalizability of findings has led Kanuk and Berenson and Linsky to call for cross-population investigations of inducement techniques effects. This paper reports one such investigation.