Incorporating person covariates and response times as collateral information to improve person and item parameter estimations


By: Shudong Wang, Hong Jiao


For decades, researchers and practitioners have made a great deal of effort to study a variety of methods to increase parameter accuracy, but only recently can researchers start focusing on improving parameter estimations by using a joint model that could incorporate RT and students information as CI. Given that many tests are currently administrated by computers and recorded RT as much other personal information is usually thrown out as test byproduct, how to use such available (may be valuable) information to improve the quality of many high stake tests has become urgent issues for states and test industries. So far in practice, the consequence of not using CI in estimation is ignored and few studies have focused on this issue; the present study attempts to provide empirical evidence on the consequence of ignoring CI on improvement parameter estimation.

See More
See the presentation

Topics: Test design

Related Topics