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

Web Survey Bibliography

Title Responsive design for mixed-mode panel data
Year 2013
Access date 11.07.2013

Many surveys today are affected by high nonresponse. This can be a serious problem to survey quality since nonresponse causes systematic error (bias) in the survey estimates. Given the decreasing trend in response rates and the corresponding increasing resources needed to achieve preset response rates, taking measures only at the estimation stage is no more sufficient to overcome this problem, nor efficient. Measures need to be taken also at the data collection stage. In this direction, different forms of responsive design have recently been proposed. The general objectives of responsive design have been formulated in Groves and Heeringa (2006). The main idea underlying this method is to intervene in the data collection process, in order to achieve an ultimate set of responding units that is “better balanced” or “more representative” than if no special effort is made. Interventions are settled by evaluating the sample properly as the data collection unfolds. To this purpose different indicators have been proposed, such as the balance and representativity indicators of the set of respondents and the distance between respondents and nonrespondents (Särndal, 2011, Schouten et al., 2009, Schouten et al., 2011, and Lundquist and
Särndal, 2012). These indicators are computable from selected auxiliary variables, which are known for the responding units as well as for the non-responding ones. By monitoring the indicators during the data collection process, it is possible to modify the original design during the course of the data collection, in order to obtain a better balanced ultimate response set. The recent existing literature presents many progresses in the development of this methodology. However, further investigations are needed in order to apply it in practice, also to different contexts. The aim of this paper is to evaluate the potentials of responsive design in the framework of mixed mode panels, where one mode is Web. The empirical application uses data from the on-going probability-based PAADEL panel. The PAADEL-Producer panel is an Italian regional panel of businesses in the agro-food sector managed at the CASI centre of Bergamo University. The recruitment of the panel was conducted in 2012 and lasted approximately three months. The first step recruitment was based on phone mode (maximum number of contact attempts five); the second step recruitment was based on the mixed mode approach (Web, phone, mail, fax). Using the
database of data collection of this research, first the progression of the estimates of a few variables is studied as the data collection unfolds. Next the balance and representativity of the panel are investigated at different steps of the recruitment. Finally, a set of experimental responsive designs based on alternative interventions in the data collection is artificially reproduced. Results are analyzed in a comparative way to evaluate the impact of this approach on the final estimates. Special attention is devoted to the bias reduction issue. Some thoughts on the consequences on the variability of the estimates are also proposed. The results obtained are promising. By way of example, Table 1 shows some of the results that will be described in the study.

Access/Direct link

Conference Homepage (abstract)

Year of publication2013
Bibliographic typeConferences, workshops, tutorials, presentations