FACE-TO-FACE, TELEPHONE AND MAIL SURVEYS - RESPONSE RATES AND RESPONSES IN A LARGE GERMAN CITY

The article focuses on the effects of different methods of data collec tion on response rates and substantive answers. The study is based on a random sample of the population of Cologne, drawn from the city regi stry. 555 persons were interviewed face to face, 362 by telephone and 308 by mail (according to the Dillman design). The largest response ra te - about 90% - was achieved by the telephone survey, followed by the mail and face to face survey (with 71% each). Noteworthy effects of t he three data collection methods on responses were not found, not even with regard to sensitive question (on self reported delinquency and w illingness to engage in it). Only in case of cannabis use there is a t endency - albeit not statistically significant - for the younger respo ndents to admit it most often in mail survey, the lowest rate is found here using telephone survey data.