Testing for the Potential for Nonresponse Bias in Sample Surveys

Publisher Website
Google Scholar
Abstract
This article proposes using a statistical hypothesis test to check for the nonresponse bias that results when there is a corrrelation between strength of opinion and reply on issue preference. This idea is developed through an example of two mail polls conducted by a Columbus, Ohio newspaper to ascertain voter preference prior to the November 1983 election.

This publication has 0 references indexed in Scilit: