Presentation paper

Examine construct validity of computerized adaptive test in K–12 assessments

2012

By: Shudong Wang, Hong Jiao

Description

The purpose of this study is to investigate the effect of missing data in computerized adaptive tests (CAT) on test construct validity. The CAT method is now becoming more popular in educational assessment. However, conducting construct validity on CAT data has unique challenges for researchers because of the nature of missing data in the CAT. Unlike linear tests in which missing mechanisms that is defined as missing type can be regarded as missing at random, the CAT algorithm determines that missing CAT is not random. The study using simulation methods examined the effect of different missing data generated from different IRT models on recovery of internal structure of tests at both item and item cluster levels. Results show it is impossible to recover the CAT test internal structure by using items as observable variables, but by parceling items and using parcels as observable variables, the test internal structure can be recovered. Parceling has the effect of over fitting models.

See More
View Asset

Related Topics