|
Abstract:
|
Technical difficulties often lead to missing item scores and hence to incomplete data for examinees on educational tests. For example, the score on an item of an online Speaking test could be missing (or, the item could be unscorable) due to poor audio quality or excessive background noise during the recording. This paper focuses on the reporting of scores for the examinees whose data are incomplete due to technical difficulties. The reporting of scores to such examinees essentially involves imputation (e.g., Rubin, 1987) of missing scores. In this presentation, data from three educational tests are used to compare the performance of six approaches for imputation of missing scores. One of the approaches, based on data mining, is the first application of its kind to the problem of imputation of missing scores. The approach based on data mining and a multiple imputation approach based on chained equations (Raghunathan et al., 2001) are found to lead to the most accurate reported scores for the examinees. A simple approach based on linear regression performed the second best overall. Several recommendations are made regarding the reporting of scores to examinees with incomplete data.
|
