| Popis: |
Violations of representativeness are a common problem in test development, especially when having to deal with combination of stratification variables (e. g. region, ethnic background, educational level ...) and given limited sample sizes in norming studies. The project aims to assess, if weighted ranking through the Harrell-Davis quantile estimator (Harrell & Davis, 1982) and using the weights in regression-based continuous norming sensu A. Lenhard et al. (2016) improves the quality of norm data. |
