in collaboration with Prof. Frauke Kreuter (University of Mannheim)
The goal of this project is to improve the data quality in web surveys using mouse movements. So far, mouse movements have mostly been analyzed using simple summary statistics and basic parametric statistical models. This project will apply and develop more advanced statistical tools in order to maximize the information drawn from the mouse movements. The development of new statistical methods is based on recent advances in functional data analysis and time series analysis. The project has an interdisciplinary set-up encompassing survey methodology and statistical methodology.
The major goals of the project in terms of survey methodology are understanding measurement error using mouse movements, relating question difficulty to mouse movements, and exploring privacy issues arising from the collection of paradata in web surveys. The major goals of the project in terms of statistical methodology are the development of regression and classification models that take (mouse) trajectories as predictor variables.