Journal of Official Statistics, Vol.12, No.1, 1996. pp. 63–83
Contact-Level Influences on Cooperation in Face-to-Face Surveys
Robert M. Groves and Mick P. Couper
Abstract:Many postsurvey adjustment procedures for unit nonresponse in surveys are based on implicit models of survey participation, attempting to estimate the response propensity of sample persons. The statistical properties of adjusted survey estimates depend on the specification of such models. To motivate such specification, a theory of face-to-face survey participation is reviewed and empirical analyses presented for one level of influences on response propensity: those influences on the decision to participate that arise during the interaction between a survey interviewer and a householder. The data suggest that traditional socio-demographic correlates of survey response rates do not predict well the outcome of individual contacts between interviewers and householders, except for the first contact. Observable behaviors of the householders, documented by interviewers after the interaction, however, are informative about the likely behavior in future contacts and about the final disposition of the sample case. Interviewers who change the content of their interactions with householders, over successive contacts, tend to achieve somewhat higher cooperation rates.
Keywords:Survey nonresponse; response propensity models; householder-interviewer interaction.
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group