In online panels it is common practice to offer incentives for study participation. The current study aimed at answering the following questions: First, how effective are different incentives in enhancing participation in web-based studies? Second, how do these incentives affect data quality? And third, how do these effects change over time?

3227 newly recruited members of an online access panel were invited to seven subsequent survey waves. One half of the participants were offered a result summary for each wave completed, the other half did not receive this offer. In addition, one out of five incentives was promised for participation in each wave. Participants were offered either no incentive, money via bank transfer, money via PayPal, loyalty points, or a donation to charity (either a predetermined charity or a selectable charity). Panelists' willingness to participate was assessed by three measures per wave: Panel attrition, response rate, and retention rate. Data quality was measured in each wave by the number of missing values, the number of words written in answers to open-ended questions, and the tendency to use mid-point categories in close-ended questions. Latent growth curve models and path models were used to analyze the longitudinal effect of the independent variables.

Independent of the manipulations, response rate decreased linearly over the seven waves. Result summary without concurrent incentives accelerated this decline. Adding an incentive to the result summary offset this negative effect, but incentives without concurrent result summary did not heighten the response rate. Donation incentives and PayPal lowered the initial response rate. The number of written words increased by a gentle quadratic slope independent of the manipulations. Either type of incentive increased this quadratic slope. Otherwise, attrition, retention, missing values, and mid-point category use were influenced by incentives merely in a sporadic manner.

The findings suggest that the use of incentives needs to be considered carefully. Panelists might perceive result summaries, Paypal and donations as a burden rather than as gratification. Straightforward payments such as bank transfer and loyalty points appear to perform better, but they need time to bring forth positive effects on panelists' willingness to participate.