ABSTRACT:
The scores of K examiners for a set of n scripts, Maximum likelihood Estimators (MLE) of true scores have been derived under the usual assumptions that the errors of evaluation and the true scores are additive and independently normally distributed. Methods for computing examiners' reliability and error variances have also been discussed using the above estimates of true scores.
Cite this article:
Prabha Rohatgi1 and Nidhi Shrivastava2 (2010). Maximum Likelihood Estimator of True Scores and Reliability of Examiners. Journal of Ravishankar University (Part-B: Science), 22(1), pp. 32-36.