Thursday, November 10
Data Quality and Measurement Error
Thu, Nov 10, 3:30 PM - 4:55 PM
Hibiscus B
Sample Design and Incentive Considerations in Pretesting and Development

Understanding Incentives for Completion: Charitable Contributions, Direct Payments, and Choice (303638)

*Catherine Wilson, 23andMe 

Keywords: data quality, engagement, response rates, compensation

Some research indicates that charitable donations can have greater value to recipients than cash rebates of an equivalent amount, suggesting that a payment to a charity on behalf of a respondent may be worth more to the respondent than the same dollar amount payment made directly to the respondent. Additionally, charitable donations may minimize potentially negative impacts of cash payments on survey taking behavior such as poorer data quality due to respondents rushing through surveys to complete them and collect payment, or reduced response rates due to the altruistic motivation to take part in research being replaced by monetary interest.

23andMe conducts research into genetics, ancestry, and health. Interviews with our research participants have shown us that they are primarily motivated to take surveys for humanitarian reasons. By reporting and updating their health histories they are contributing to research toward genetic discovery that they believe is valuable and beneficial to both themselves and others.

This paper reports the findings of an experiment we embarked on at the start of 2015, studying the impact of offering different kinds of compensation for completion of a nine-month longitudinal study. Each of 20,000 research participants in the study were randomly assigned to 1 of 6 groups, which varied in whether they were offered direct money compensation, charitable contributions of different dollar values, or a choice between each. The results of this experiment will include an analysis of the impact of different compensation offerings on response rates, data quality and respondent effort, as well as willingness to take subsequent health surveys for which no compensation is offered. From this we hope to learn about and share with other researchers best practices in compensating participants so that completions, quality, and long-term participant engagement are optimized.