Going online with the health interview survey? Assessing the effect of data collection mode on participation, measurements and costs in a Belgian context

National health interview surveys still have a key role in our health information system. These surveys are traditionally organized by means of a face-to-face interview, often in combination with a self-administered paper-and-pencil questionnaire for the most sensitive topics. This approach is being questioned due to some challenges (e.g. decreasing response rates, high financial costs and logistical burdens…) and the advantages associated with a new mode, namely web (e.g. cost advantage, ease of implementation…). Moreover, speedy internet connections and low internet access costs have contributed to high internet penetration rates (85% of the European households had access in 2016), making the web even more appealing for surveys. The web mode can be implemented as a single mode of data collection or within a mixed-mode design. The latter means that some respondents fill in the questionnaire via the web mode while others use the traditional mode (e.g. because of no internet access). The use of other modes of data collection or mixed-mode data collection will not only have implications for the organization of the fieldwork and for the costs, but can also lead to differences in the quality of the collected data. For example, it can affect the response rates (and the risk of non-response errors) and the measurement of health indicators (and the risk of measurement errors). In this dissertation, the effect of online data collection was assessed in the context of the Belgian health interview survey. The following aspects were, amongst others, evaluated: What is the effect of using different modes of data collection on the estimates of health indicators? What is the effect of applying different modes of data collection on the unit and item response rates? What is the difference in costs between different modes of data collection? This was done by organizing three studies (one in a convenience sample, the others in a random sample drawn from the general population). By means of this PhD project, the existing knowledge on the use of web and mixed-mode data collection in the framework of health interview surveys expanded. Moreover, the dissertation includes some general recommendations in order to meet the current challenges associated with the organization of health interview surveys.

Publication year:2020

Keywords:Doctoral thesis

Accessibility:Open