摘要：This paper describes two postal surveys conducted in New Providence between September 2001 and February 2002. These surveys, using the telephone directory as the sampling frame, were specifically designed to investigate the response rate to postal surveys in New Providence. Each survey included a one-page form sent to 250 persons resulting in a response rate of about 20%. Each survey form contained one question that had also appeared in The Bahamas 2000 census. Overall response to this question differed significantly between the respondents to each survey and the respondents to the Census. These results suggest the presence of sources of bias relevant to the sampling frame, socio-economic status and self-selection. The implications of these results for using the postal survey as a research tool are discussed.
