Journal article
Classification Consistency and Accuracy With Atypical Score Distributions
Journal of educational measurement, Vol.57(2), pp.286-310
06/01/2020
DOI: 10.1111/jedm.12250
Abstract
The current study aims to evaluate the performance of three non-IRT procedures (i.e., normal approximation, Livingston-Lewis, and compound multinomial) for estimating classification indices when the observed score distribution shows atypical patterns: (a) bimodality, (b) structural (i.e., systematic) bumpiness, or (c) structural zeros (i.e., no frequencies). Under a bimodal distribution, the normal approximation procedure produced substantially large bias. For a distribution with structural bumpiness, the compound multinomial procedure tended to introduce larger bias. Under a distribution with structural zeroes, the relative performance of selected estimation procedures depended on cut score location and the sample-size conditions. In general, the differences in estimation errors among the three procedures were not substantially large.
Details
- Title: Subtitle
- Classification Consistency and Accuracy With Atypical Score Distributions
- Creators
- Stella Y. Kim - Univ N Carolina, 9201 Univ City Blvd, Charlotte, NC 28223 USAWon-Chan Lee - Univ Iowa, 210 E Lindquist Ctr, Iowa City, IA 52242 USA
- Resource Type
- Journal article
- Publication Details
- Journal of educational measurement, Vol.57(2), pp.286-310
- Publisher
- Wiley
- DOI
- 10.1111/jedm.12250
- ISSN
- 0022-0655
- eISSN
- 1745-3984
- Number of pages
- 25
- Language
- English
- Date published
- 06/01/2020
- Academic Unit
- Psychological and Quantitative Foundations
- Record Identifier
- 9984371299802771
Metrics
10 Record Views