Abstract

Our research demonstrates how supervised learning models categorize answers to an open-ended question on survey participation motivation. The GESIS Panel, a German access panel with 5,400 panelists, has collected around 25,000 answers since 2014. We differentiate between extrinsic and intrinsic motives and examine changes in motivation over years of panel participation. By using a coding scheme and a Support Vector Machine algorithm, we automated the categorization of answers.
Results show that panelists are motivated by interest, incentives, recognition, and a desire to help the scientific community and society.