Effects of Model Specification in Subscale Score Estimation in International Large Scale Assessments (PhD project)
How best can we estimate overall- and subscale-scores in international large scale assessments? (Copyright: Colourbox)
About the project
International educational assessments are a useful tool in monitoring performance and progress of participating countries. However, most operational methods currently used to obtain achievement scores were developed under the assumption that they are universally applicable. That is to say, these approaches assume that “one size fits all.”
However, these scoring methods may be sensitive to diversity in participating countries. Therefore, this project intends to evaluate the interactive effect of test- and model-specification on the process of estimating overall- and subscale population scores when diverse populations are tested. This will be crucial in knowing how well current scoring methods perform in diverse populations; and, ultimately, provide suggestions as to how they may be improved.
This project is located at the Centre for Educational Measurement at the University of Oslo (CEMO). It runs between November 2016 and November 2020. The project will comprise of a series of simulation studies and two real data examples, which will be differentiated by their complexity in sampling designs.
Parent research project
The project is funded by the Research Council of Norway (Project number: 255246)