Tuesday 16th July
Wednesday 17th July
Thursday 18th July
Friday 19th July
Download the conference book
Mixed Mode Surveys - Reports from the Field Work 2 |
|
| Chair | Mr Patrick Schmich (Robert Koch-Institute) |
Primary data collection among ethnic minorities is generally cost-intensive, as interviewers and questionnaires often have to be bilingual and respondents are often hard to reach. In addition, researchers face the problem of comparatively small (ethnic-)subgroups and are in need of high response rates. Our motivation was to find a cost-saving mode of data collection that would allow us to invest resources in cash incentives to increase participation rates. We are going to begin by introducing theoretical considerations regarding the use of different forms of incentives. Thus, a "double incentive" strategy consisting of an unconditional and a conditional part shall be discussed. The logic of this strategy is to simultaneously increase the costs of non-participation and the benefits of participation for interviewees. Empirical evidence is presented on the basis of a mail survey experiment conducted among young German-Turkish adults (N=500) which are split into five groups; a control group, two unconditional only, and two double incentive groups using cash incentives up to 15 Euro. Since we had access to two geo-data items on the street-level we have some information about non-respondents. Our findings suggest that the double incentive strategy has positive effects on response rates. Furthermore, in accordance with prior findings incentives seem to reduce sample bias. Obviously, they lead to a decrease in sampling bias of respondents from disadvantaged subgroups. We find no evidence with respect to item non-response or willingness to participate in another interview throughout the groups.
The current economic situation has put a strain on research funding and both survey funders and survey agencies need to find innovative ways of delivering high quality research for a lower cost. Mixing modes is seen by many as one of the ways to achieve this aim. It is expected that starting with a cheaper mode and following it up with a more expensive mode reduces the costs while not affecting or possibly improving the response rates. There is not much research evidence to show that this is the case, especially in the context of mixing web and face-to-face, the two extremes on the cost spectrum. This paper aims at shedding some new light to this area. We will use evidence from a mixed modes experiment that was included in the Understanding Society Innovation Panel waves 5 and 6 in the UK. The Innovation Panel sample was randomly split to an experimental group that was first invited to take part in the web survey and in case of non-response followed up face-to-face, and a control group that was only approached face-to-face. This offers a unique opportunity to disentangle the effect that mixed modes approach has on response, measurement and cost. In our paper will focus on two these - the response and cost, showing that mixing modes can depress rather than improve response and that the cost implications are not as univocal as might be expected.
Up to and including 2001, the Italian population and housing census was conducted with the conventional methodology of complete field enumeration (so-called "door-to-door"). Census forms were delivered and collected by enumerators and self-filled in by respondents. All information was collected and processed on a complete basis (without making use of any sampling techniques) while the same economic, human and organisational resources were allocated to every household. The need of reducing municipalities' workload and the burden on respondents has called for the use of new data collection techniques and new territorial instruments meant to improve coverage and quality of the enumeration. The once "door-to-door" census has become a register-supported census, implemented by means of questionnaires' mail out to the households enrolled in the municipal population registers. Self-completed questionnaires were then collected by a multi-mode data collection system allowing households to choose the way in which they preferred to complete and return the questionnaire: online, using the password provided with the questionnaire; at any post office in Italy; at one of the municipality census collection centres, at which specialist assistance for the completion of questionnaires was also available; directly to a municipality enumerator, in service for the completion of census operations. The innovations designed for 2011 Italian census are not enough to achieve a stable and enduring balance between census costs and benefit. For these reasons the development of a completely different approach seems necessary.
The German Health Update (GEDA) is a periodic cross-sectional telephone survey, representative for the population in Germany and delivers data on a variety of health indicators. As a matter of fact, conducting telephone surveys means researchers are increasingly confronted with challenges consisting mainly of rising costs and declining response rates. Facing these problems a methodological study was carried out in 2012 (N=10.080; age: 18-79) using different types of mixed-mode designs, e.g. sequential and parallel mixed-mode design. Subjects were randomly selected from official registers of six local municipalities, ensuring representation of rural and urban as well as eastern and western municipalities. Additionally the sample was stratified according to age and gender. The study used paper and pencil interviews (PAPI), web interviews (CAWI) and computer assisted telephone Interviews (CATI) to collect data. Subjects were contacted up to three times by mail.
Although the results show that administering surveys using mixed-mode design are more complex than surveys using single-mode design they can be considered as a promising alternative to deal with the challenges population-based surveys are confronted with today.
Comparing sequential and parallel mixed-mode designs the following presentation will report on the challenges related to organizing the field work and will describe experiences and lessons learned from the project. Furthermore we would like to discuss the consequences of the results in terms of prospective study designs for the German Health Update.