Improving predictions of response propensities for effective Adaptive Survey Design (ASD)

Adaptive survey designs are often optimized using historic survey data. Response propensities for a range of relevant socio-economic background characteristics are estimated and form the input to mathematical optimization problems. Apart from being subject to sampling variation, propensities also change over time.
This dissertation investigates the utility of Bayesian analysis and applies various approaches to the optimization of desigsn accounting for uncertainty and time change.
Wu, S. (2023). Improving predictions of response propensities for effective Adaptive Survey Design (ASD). Dissertation, Utrecht University, doi:10.33540/1791.
Downloads
- PDF - Dissertation Wu