Many contexts for test equating, such as for university selection, are complex with different students sitting different combinations of tests. This thesis shows that using only each student's total score on each test, the Rasch model of modern test theory can be applied to equate multiple tests simultaneously, even when not all students have scores on all tests. Two advantages of the approach are that no ad-hoc preliminary steps, such as smoothing of scores, are required, and that the equated scores can be interpreted in the same way that a measurement is interpreted in physical science.
|Qualification||Doctor of Philosophy|
|Award date||21 Jan 2020|
|Publication status||Unpublished - 2020|