Eksplorasi Faktor-Faktor Korelasi Antara Classical Test Theory dan Item Response Theory
| Main Authors: | Al Fajri, Abdurrahman Nadhif, Widhiarso, Wahyu |
|---|---|
| Format: | Article PeerReviewed application/pdf |
| Bahasa: | eng |
| Terbitan: |
Unit Pengembangan Alat Ukur (UPAP)
, 2019
|
| Subjects: | |
| Online Access: |
https://repository.ugm.ac.id/276962/1/Alfajri%20-%20Widhiarso%202019%20-%20Penyekoran%20IRTi.pdf https://repository.ugm.ac.id/276962/ |
Daftar Isi:
- The main purpose of assessing abilities in the educational environment is to make various decisions based on the results of the exam. A scoring method must have the ability to produce high precision scores. To achieve this, test administrators usually use one of two popular scoring methods; Classical Test Theory and Item Response Theory. These Theories have their own advantages and disadvantages. IRT have high scoring precision, but the scoring process is difficult. On the other hand, CTT scoring comes with easy scoring processes but lacks precision. This study explores the factors that can lead to different scoring results through the CTT and IRT approaches while examining how far the scores produced by the two methods are related. The factors involved in this study were variations in item parameters, number of samples, number of items, and distribution of participants' abilities. Analysis of 216 factor combinations using one-way ANOVA found that there were significant differences in the correlation between combinations of factors. Meanwhile, the highest correlation between the two types of scores was found in conditions with irradiated item parameters, 1000 participants, 60 items, and normally distributed participants’ abilities. Further research is needed to explore other factors beyond this research.