PT  - JOURNAL ARTICLE
AU  - Richard Colombo
TI  - A Model for Diagnosing and Reducing Non-response Bias
AID  - 10.2501/JAR-40-1-2-85-93
DP  - 2000 Jan 01
TA  - Journal of Advertising Research
PG  - 85--93
VI  - 40
IP  - 1-2
4099  - http://www.journalofadvertisingresearch.com/content/40/1-2/85.short
4100  - http://www.journalofadvertisingresearch.com/content/40/1-2/85.full
SO  - J Advert Res2000 Jan 01; 40
AB  - Perhaps the first question people ask when judging the quality of a sample survey is, ‘What is the sample size?’ Just as important, but one less often asked, is the question ‘What is the response rate?’. Response rate is important in assessing the non-response bias, which is independent of sample size and usually far exceeds sampling error as a source of uncertainty. This discussion paper looks at current ‘callback’ and weighting methods, and proposes a model to reduce response rate error.