webinar register page

Online Calibration Designs and Methods for Multidimensional Computerized Adaptive Testing
Ping Chen, Beijing Normal University
Chun Wang, University of Washington

Computerized adaptive testing (CAT), as the earliest form of instantiation of artificial intelligence in assessment design, has been widely used in the fields of psychological testing, educational evaluation, personnel selection, and medical diagnosis. The continuous implementation and sustainable use of CAT rely heavily on periodical update and replenishment of the item bank. While generating large set of items by humans or machines is an important research topic by itself, calibrating those item parameters efficiently and accurately is another research topic, known as online calibration, that will be the focus of this webinar. Online calibration includes both the design of assigning items to individuals and the algorithmic techniques to estimate item parameters. With the increasing trend of measuring complex constructs such as higher-order thinking in education or multifaceted function in health domains, multidimensional CAT is needed to optimize efficiency. This webinar will focus on online calibration in multidimensional CAT.

First, we will briefly describe the background knowledge of online calibration, including its definition, implementation steps, and data structure examples; second, we introduce the two important aspects of online calibration (i.e., online calibration method and online calibration design), followed by our representative research work and progress in this area; then, we reveal the relative advantages, disadvantages, and applicability of various methods/designs via comprehensive simulation studies; finally, we prospect the future research directions in this field.

May 12, 2022 11:00 AM in Eastern Time (US and Canada)

Webinar is over, you cannot register now. If you have any questions, please contact Webinar host: Susan Rees.