Examining the Stability of Angoff's Delta Item Bias Statistic Using the Bootstrap

Publisher Website
Google Scholar
Abstract
Item bias indices are frequently used in the process of screening items. Although many bias indices exist and much has been written about the agreement between indices, relatively little has been done to investigate the stability of a particular index across comparable samples. Using the bootstrap methodology, this study investigated the stability of the Angoff delta item bias index. The results support this use of the new methodology, but suggest that sampling error in estimating the statistic needs to be closely monitored.

This publication has 4 references indexed in Scilit: