Exploring and Correcting for Nonresponse Bias Using Follow-ups of Non Respondents

This article is primarily intended as a critique of "The Use of Potential Nonrespondents for Studying Nonresponse Bias," a paper prepared for this journal by R. A. Ellis, C. M. Endo, and J. M. Armer (Ellis et al., 1970). The author assumes that the purpose of sample surveys is to estimate population parameter values while minimizing the effect of nonresponse bias. It is argued that data from successive cumulative response waves may be used to compute corrected parameter values by using a linear regression. Most researchers have learned from experience that it is virtually impossible to obtain replies from every person in large sample surveys. Ellis et al. (1970) emphasize the crucial importance of minimizing error due to nonresponse. When respondents differ from nonrespondents statistics based on responses alone often do not validly depict the population investigated and may result in predictions which are inaccurate,