Tuesday 14th July      Wednesday 15th July      Thursday 16th July      Friday 17th July     

Thursday 16th July, 09:00 - 10:30 Room: HT-105

Push2Web and Nudge2Web: Combining Mail and Online Survey Modes to Reduce Survey Errors and Survey Cost
Convenor	Professor Henning Best (University of Wuerzburg )
Coordinator 1	Professor Michael Bosnjak (GESIS - Leibniz Institute for the Social Sciences)
Coordinator 2	Mrs Tanja Dannwolf (GESIS - Leibniz Institute for the Social Sciences)

Session Details

The sharp decline in telephone survey response rates has strongly increased the need for alternative ways for surveying the general public. Web surveys with offline recruiting are a potentially promising option, but at the moment internet use and coverage is still limited. As a possible solution to this problem, Dillman and Colleagues have recently suggested using a sequential mixed-mode design combining mail and web survey modes (see e.g. Messer & Dillman 2011). The key idea is to initially contacting the respondents by mail, ´pushing´ them to the web mode (Push2Web) and finally providing a paper questionnaire as a last resort to avoid unit nonresponse. The type of mode the respondents are expected to participate in is under the control of the researcher. Another conceivable strategy is to leave the mode choice to the respondents, and using various strategies ´nudging´ them to the web mode (Nudge2Web).

While these approaches appear promising, important questions on survey errors and biases (coverage, sampling, nonresponse, measurement) and survey costs, as well as questions of practical survey management remain understudied.

This session is aimed at bringing together pertinent research and recent experiences on combining offline recruitment and mixed-mode (especially mail+web) surveys from Europe, and encourages submissions about the impact of various Push2Web or Nudge2Web strategies on survey errors and costs.

Paper Details

1. Measurement and Cost effects of Pushing Household Survey Respondents to the Web for Surveys of Electricity and Gas Customers in the United States
Professor Don Dillman (Washington State University)
Dr Michael Sullivan (Nexant, Inc.)
Ms Caren Leong (Nexant, Inc.)

This paper reviews 19 U.S. surveys of electric and gas utility customers conducted by Nexant, Inc. in which variants of a “push2web” methodology were used. Mail and email contacts, as well as the possibility of responding by web or mail were tested in various combinations in order to evaluate the response quality, costs, and how well the demographics of the completed samples matched the population is from which those samples were drawn. This paper significantly extends research on how to improve web response rates and data quality by combining multiple modes of contact as well as response.

2. Combining web and mail sequentially – error and cost advantages in the Swiss context
Professor Caroline Roberts (University of Lausanne)
Dr Michèle Ernst Stähli (FORS - Swiss Centre of Expertise in the Social Sciences)

Empirical evidence confirming the benefits of mixing web and mail in the European context is still relatively limited. In this paper we present results from mode comparison experiments carried out in Switzerland. We focus on the question of whether using mail follow-ups to contact nonrespondents to a web survey (for which respondents were recruited off-line) helps to improve sample representativeness, and to reduce nonresponse bias. We combine two sources of data to address these questions: auxiliary data from the register-based sampling frame, and data from a nonresponse follow-up survey.

3. Sequential Mixed Mode Experiment in the US National Monitoring the Future Study
Dr Megan Patrick (University of Michigan)
Dr Mick Couper (University of Michigan)
Dr John Schulenberg (University of Michigan)

Monitoring the Future (MTF) is a US study of drug use. Each year, a nationally representative sample of 12th-graders (age 18) is surveyed and a subsample is followed longitudinally using paper questionnaires. To investigate a web-based follow-up option, investigators conducted a sequential mixed mode experiment in 2014 with various strategies to promote (“push”) either postal mail or web-based response. Participants (N=4,941) were randomly assigned to three conditions: Mail Push, Web Push, or Web Push + Email correspondence. The presentation will focus on response rates by condition and differences in data quality and responses by mode.

4. Mixed-Mode strategies in Business Surveys: Experimental Results
Professor Henning Best (University of Wuerzburg)

This paper compares different recruitment strategies for a mixed-mode-survey (mail and web) of small and medium enterprises. We randomly assigne participants to experimental groups, varying the intensity of pushing them to participation via web. In the first invitation, 50% of the participants receive a paper questionnaire, and 50% of that group an additional web link. The no-paper group receives a web link only. In the second and third wave of invitations, the experimental design is continued. Based on this design we can empirically compare push2web and nudge2web strategies in business surveys.