Survey of Computerized Adaptive Testing: A Machine Learning Perspective

arXiv:2404.00712v1 Announce Type: new
Abstract: Computerized Adaptive Testing (CAT) provides an efficient and tailored method for assessing the proficiency of examinees, by dynamically adjusting test questions based on their performance. Widely adopted across diverse fields like education, healthcare, sports, and sociology, CAT has revolutionized testing practices. While traditional methods rely on psychometrics and statistics, the increasing complexity of large-scale testing has spurred the integration of machine learning techniques. This paper aims to provide a machine learning-focused survey on CAT, presenting …

abstract adjusting arxiv cs.ai cs.cy cs.ir cs.lg diverse education fields healthcare machine machine learning performance perspective practices questions sociology sports survey test testing type