Test Theory and Artificial Intelligence Based Technologies for Performance Assessment

Abstract

In various assessment contexts including entrance examinations, educational assessments, and personnel appraisals, performance assessment has attracted much attention to measure examinees’ higher order abilities. Nevertheless, low assessment reliability and high costs of scoring are regarded as persistent difficulties hindering performance assessment. To resolve these shortcomings, item response theory models that incorporate rater and task characteristic parameters and automated essay scoring methods have been proposed recently. This paper introduces state-of-the-art topics for these technologies.