Journal article
Procedures for Extending Item Bias Detection Techniques to Polytomously Scored Items
Applied measurement in education, Vol.6(1), pp.1-19
01/01/1993
DOI: 10.1207/s15324818ame0601_1
Abstract
The recent increased emphasis on various types of performance assessments raises the question of what is the effect of such assessments on various population subgroups. This study was designed to suggest methodology for several statistical procedures to detect performance assessment items that function differently for two subgroups of examinees. Three different statistics are proposed. Each uses an external criterion (e.g., performance on a multiple-choice test) as a conditioning measure. All three methods being compared are alternative ways of combining data from 2 x k tables. Computer simulations were conducted to study the behavior of these statistics in the detection of items exhibiting varying degrees of differential item functioning. The t statistic-based methods appeared to be somewhat more powerful, and one of these two statistics also demonstrated better control of Type I error under certain conditions. © 1993, Taylor & Francis Group, LLC. All rights reserved.
Details
- Title: Subtitle
- Procedures for Extending Item Bias Detection Techniques to Polytomously Scored Items
- Creators
- Catherine WelchH Hoover
- Resource Type
- Journal article
- Publication Details
- Applied measurement in education, Vol.6(1), pp.1-19
- Publisher
- Lawrence Erlbaum Associates
- DOI
- 10.1207/s15324818ame0601_1
- ISSN
- 0895-7347
- eISSN
- 1532-4818
- Language
- English
- Date published
- 01/01/1993
- Academic Unit
- Psychological and Quantitative Foundations
- Record Identifier
- 9984371261202771
Metrics
2 Record Views