Item Characteristics of National Examination Council’s Economics Multiple-Choice Items: An Item Response Theory Exploration
The study explored the IRT parameter estimates of Economics multiple-choice items using the 1, 2 and 3 parameter logistic models. The study adopted the explorative research design with a sample size of 1500 senior secondary school III Economics students’ selected using multi-stage sampling procedure...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Atatürk University
2024-04-01
|
| Series: | Journal of Psychometric Research |
| Subjects: | |
| Online Access: | https://dergipark.org.tr/tr/download/article-file/3770688 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The study explored the IRT parameter estimates of Economics multiple-choice items using the 1, 2 and 3 parameter logistic models. The study adopted the explorative research design with a sample size of 1500 senior secondary school III Economics students’ selected using multi-stage sampling procedure. The Economics Achievement Tests (EAT) and students’ responses as contained in the optical mark reader (OMR) serve as instruments. Data collected was coded and analysed using Mirt package in R statistical software for item parameter calibrations. The results showed that the discrimination index estimated using the 2 Parameter Logistic (PL) and 3PL models indicated that 28 items and 25 items respectively are poor items while 32 items and 35 items are considered good items respectively. Also, the difficulty index estimated using the 1, 2 and 3 PL models shows that 23 items, 25 items and 35 items respectively are easy items, 35 items, 33 items and 23 items are moderately difficult items while 2 items are considered difficult items. Furthermore, the results of the 3PL model shows that only 9 items are considered to be vulnerable to guessing with 51 items not vulnerable to guessing. The study concluded that the IRT psychometric estimates of NECO Economics multiple-choice items possessed moderately difficult items with an average discrimination indices and items majorly found not vulnerable to guessing. It therefore recommended that test experts and examination bodies should regularly consider the use of IRT psychometric estimations to evaluate item parameters for quality check of test items. |
|---|---|
| ISSN: | 3023-4646 |