Title: Accurate Assessment via Process Data
Abstract: Accurate assessment of an examinee's ability is the key task of educational tests. Scoring based on final responses to test problems is the standard. As the infrastructure advances, substantially more information is observed. One of such instances is the process data, i.e., the computer logged events of an examinee's detailed interactions in pursuit of solving a problem. We show both theoretically and with simulated and empirical data that appropriately including such information in scoring, through a Rao-Blackwellization step, can substantially improve the precision of ability measurement.
