Explanatory Item Response Models for Polytomous Item Responses
Item response theory is a widely used framework for thedesign, scoring, and scaling of measurement instruments. Item response modelsare typically used for dichotomously scored questions that have only two scorepoints (e.g., multiple-choice items). However, given the increasing use ofinstruments that...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
izzet kara
2019-07-01
|
| Series: | International Journal of Assessment Tools in Education |
| Subjects: | |
| Online Access: | https://dergipark.org.tr/tr/download/article-file/716984 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Item response theory is a widely used framework for thedesign, scoring, and scaling of measurement instruments. Item response modelsare typically used for dichotomously scored questions that have only two scorepoints (e.g., multiple-choice items). However, given the increasing use ofinstruments that include questions with multiple response categories, such assurveys, questionnaires, and psychological scales, polytomous item responsemodels are becoming more utilized in education and psychology. This study aimsto demonstrate the application of explanatory item response models to polytomousitem responses in order to explain common variability in item clusters, persongroups, and interactions between item clusters and person groups. Explanatoryforms of several polytomous item response models – such as Partial Credit Modeland Rating Scale Model – are demonstrated and the estimation procedures ofthese models are explained. Findings of this study suggest that explanatoryitem response models can be more robust and parsimonious than traditional itemresponse models for polytomous data where items and persons share common characteristics.Explanatory polytomous item response models can provide more information aboutresponse patterns in item responses by estimating fewer item parameters. |
|---|---|
| ISSN: | 2148-7456 |