Notice: the WebSM website has not been updated since the beginning of 2018.

Web Survey Bibliography

Title Item Nonresponse in a Mixed-Mode Household Travel Survey
Author Pena, R. J.; Wilbur, M.; Kolenikov, S.
Source The American Association for Public Opinion Research (AAPOR) 70th Annual Conference, 2015
Year 2015
Access date 07.07.2015
Abstract

Item nonresponse is a common indicator of data quality. Demands for data quality are frequently balanced against greater project efficiencies and cost savings considerations in mixed-mode household surveys. To the extent item nonresponse can be relied upon to evaluate data quality, developing strategies for item nonresponse minimization will be desirable. Early research on item nonresponse in surveys leaned on cognitive and social psychological frameworks to explain effects associated with face-to-face interviewing, respondent burden, motivation to comply, topic salience, perceived threats, and privacy needs. Technological advances in data collection have obviated the need for some item nonresponse minimization practices while also introducing entirely new challenges. Drawing on data from a recently completed household travel survey conducted in the MidSouth region of the United States, we uncover both common and unique patterns of item nonresponse. In this presentation, we analyze effects on item nonresponse from: a) data collection mode; b) question content type; c) differential incentives; d) household characteristics; and, e) proxy reporting procedures in household travel surveys. Preliminary findings suggest: a) data quality concerns were greater for interviews completed exclusively on the web compared to mixed web+paper/phone+paper/phone+web modes; b) item nonresponse rates were higher for personal work/school location reporting in comparison to other destination reporting; c) promised incentive structure yielded mixed benefits; d) respondents who used initials rather than full household member names had significantly higher item nonresponse rates to personal questions like income, household member age, education, place of employment; and, e) self-reporting was more reliable than proxy reporting overall. We discuss strategies to reduce item nonresponse including obtaining complete recontact information from respondents, assisting the primary respondent in garnering cooperation among all household members, and applying questionnaire design techniques aimed at lessening respondent burden and increasing trust in sponsorship.

Print

Web survey bibliography (4086)

Page:
Page: