Personalization, respondent anonymity, and response distortion in mail surveys.

Investigated whether the level of personalization expressed in an advance notice and cover letter distorted the reporting of common demographic information in a mail survey. A factorial design with 2 levels of advance notice (telephone call, mimeographed form letter) and 2 levels of cover letter (personalized, form) was employed. Ss were 276 married females (representing a 42% response rate) for whom 7 items of demographic information were available from an independent source (credit application). Comparison of mail-survey and credit-application responses revealed that distortion in reporting family income and occupation of wife was related to the use of personalized cover letters. Implications of the findings for behavioral scientists who depend on mail surveys for data collection are considered. (16 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)