Relevance & research question: Positive effects of incentives in longitudinal surveys for initial participation are well-documented in the literature. Introducing or increasing an incentive amount on panels has also received scholarly attention. However, increased usage of incentives raises concerns about creating respondents' expectations of being paid for survey participation. Our research goal is to answer the question whether respondents form incentive expectations after having participated in several monthly surveys or whether they are motivated by their experience within the panel. The latter would allow saving on incentives in later waves.
Data and method: We analyze data from the offline-recruited probability-based online panel conducted by GESIS. Starting in February 2011, respondents were surveyed every month for 8 months. The panel implemented differential incentives: the amount was varied experimentally during recruitment, comprising groups of 0, 2, 5 and 10 EUR. Additionally, we tested the effect of a one-time 20 EUR bonus payment for responding to 8 questionnaires in some groups. At the end of the 8th online survey we requested an agreement for further participation. Those respondents, who agreed, received an invitation to the 9th survey, for which the incentive was changed to 2 EUR (increase for 0-EUR-group, no change for 2 EUR group and a decrease for 5- and 10-EUR groups).
Results: First results indicate that incentive manipulations do not affect completion rates of the wave where incentive reduction or increase was implemented. The previous response history is one of the most significant predictors for participation in the 9th wave. Measures of survey enjoyment from the previous wave can only partly predict longitudinal response.
Added value: We report reassuring findings that respondents are not affected by possible expectations of a specific incentive. This implication may be of interest if faced with budget constraints, which make increase of incentives not affordable or even force to consider a reduction of future incentives. As large-scale panel surveys generally do not employ incentive manipulations for ethical or logistical reasons, our results can have a practical value in panel management strategies.