Mode and incentive effects in an individual register frame based Swiss election study

We analyze the effects of a randomized telephone / web mode experiment in a Swiss election survey. The web sample was further randomly divided into a group who received an unconditional 20 Sfr. postal cheque and a group without incentive. The sample of each of the three resulting designs is representative of Swiss citizens with the right to vote and was drawn at random from a register of individuals. This register guarantees almost full coverage and includes basic socio-demographic characteristics. We examine various survey quality measures in all three designs: response rates, sample composition, item nonresponse, straightlining, primacy, recency, and cost. While the response rates are comparable in the telephone and the web with incentive designs, in the web mode without incentive it amounts to about half of the one with incentive. The incentive is found to strongly increase the response rate especially for young people, while the effect on older people is limited. While young people are underrepresented in the telephone survey, and overrepresented in the web surveys, older people are underrepresented in all designs, but even more so in the web surveys. Item nonresponse occurs to a higher extent in the telephone survey. There are few differences with respect to straightlining, primacy and recency effects. Regarding cost, the web survey is much less costly, even with an incentive. Based on these findings we tend to consider web with incentive as a real alternative to the telephone mode usually used in the survey examined, especially when taking into account its reduced costs, and its comparatively high survey data quality. In the long term, while the increasing coverage problems of telephone surveys speak against this mode, the opposite is true for web surveys. This might eventually solve the remaining problem of reaching older people with web surveys.