The impact of ChatGPT service on students’ performance: Moderated by training

This research paper aims to test the effects of ChatGPT on students’ performance while using training to moderate this effect. The current paper uses a quantitative, descriptive, cause-effect approach. A cross-sectional sampling approach was used to collect the data online from 117 students in three Jordanian universities (Princess Sumaya University, University of Jordan, and German Jordanian University) by using a survey questionnaire. Data has been tested for its validity and reliability before testing hypotheses. The results indicated that the students agreed on the importance of ChatGPT (ease of use, accuracy, and plagiarism), however, most of the respondents did not agree on the importance of training on ChatGPT and they say it is easy and does not need training. The results also show that there are significant correlations among ChatGPT dimensions (ease of use, accuracy, and plagiarism). However, there is a significant correlation between training and plagiarism only, and there is an insignificance between training and both ease of use and accuracy, which supports the respondents' viewpoint that the training is not important. Finally, findings indicate that there is a significant strong correlation between all other variables (ease of Use, accuracy, and plagiarism) and students' performance, and a weak relationship with training. Finally, results show that there is a significant impact of ChatGPT (Accuracy, ease of use, and plagiarism) on students’ performance, where plagiarism has rated the highest significant effect, then accuracy, while ease of use has an insignificant effect. Moreover, results demonstrated that training has an insignificant moderation effect between ChatGPT and students’ performance.