Journal of Official Statistics, Vol.4, No.2, 1988. pp. 155–172
Miscellanea, Self/Proxy Response Status and Survey Response Quality, A Review of the Literature
Jeffrey C. Moore
Abstract:Three decades of research have not produced conclusive evidence of consistent response bias or response error variance differences due to self/proxy response status. The net nonresponse effect may also be close to zero due to compensating effects for the various components of nonresponse. The main cause of the lack of evidence is the methodological shortcomings of much of the research which purports to address the self/proxy issue. In addition, the few methodologically sound studies – most importantly, those which control potential self-selection biases, and whose subject matter makes the self/proxy distinction appropriate – in general have produced no effects or conflicting effects (or, in the case of nonresponse, compensating effects). However, lack of convincing evidence of quality differences is not synonymous with convincing evidence of no quality differences. Until more data are gathered, the conclusion that self and proxy survey responses are of equivalent quality must remain tentative.
Keywords:Self/proxy response; response quality; response bias; response error variance; nonresponse; survey design.
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group