The transition of an in- person panel survey to a web-mail mixed mode design
Jul 25, 13:45
This presentation discusses some of the issues in transitioning an in-person longitudinal survey to a self-administered, web/paper mode with in-person nonresponse followup. It reports on a three-year research effort conducted for the National Longitudinal Study of Adolescent to Adult Health (Add Health) that tested a range of design options including incentives, mail contact protocols and questionnaire length. The Add Health follows a nationally representative sample of about 19,800 adolescents selected from grades 7-12 in the 1994-95 school year. The study is currently completing its fifth wave of interviews; Wave IV concluded in 2008. Waves I - IV were conducted by in-person administration using a combination of computer assisted personal interviewing (CAPI) and computer assisted self-interviewing (CASI). To reduce costs, Wave V transitioned to a mixed-mode, two-phase survey design. Phase 1 is a mail survey conducted by computer assisted web interviewing (CAWI) and paper and pencil interviewing (PAPI). Phase 2 selects a subsample of nonrespondents for CAPI/CASI field interviewing and allocates the remaining nonrespondents to computer assisted telephone interviewing (CATI) followup using an abbreviated questionnaire. The Wave V sample is being fielded as four random subsamples over an approximately three-year period. Experiments are being conducted in all four-subsamples to inform the design of subsequent data collections. These experiments focus on incentive structure, the questionnaire design and the data collection protocols. Nonresponse followup is conducted in-person on a random sample basis. This paper describes the Add Health Wave V survey design, provides some results of the experiments and discusses the implications for general in-person surveys transitioning to web or paper self-administered modes.