A Study on the Analysis of the AI Educational Contents Usability Test Appropriateness by Rasch Rating Scale Modeling
Gyun Heo · Journal of Fisheries and Marine Sciences Education · 2022
The purpose of this study is to analyze the appropriateness of item, item difficulty, rating scale, and compare people"s ability to item difficulty with the AI Educational Contents Usability Test(AI-ECUT) scale through applying Rasch Model based on the item response theory. The data are 105 students responding survey after attending to usability test. For data processing, frequency analysis, factor analysis, and Rasch model analysis are performed using jMetrk, Winsteps, and SPSS program. The results are as follows. First, it is found that the AI-ECUT scale is found to satisfy one-dimensional fitness. Second, it is found that appropriate to change the rating scale category from 5 points to 4 points Likert scale. Third, as a result of examining the item relevance of the AI-ECUT scale, the Q41 item is found to be inappropriate and needs to revise. Forth, there are lots of lower difficulty levels items compared to higher and middle difficulty levels items.