Exploring ChatGPT-Generated Assessment Scripts of Probability and Engineering Statistics from Bloom’s Taxonomy

Subject Instructors, class teachers, and educational practitioners always devote much time to preparing assessment scripts, suggested solutions, and marking schemes such as mid-term and final examination scripts for assessing students’ learning and performance as well as measuring their achievements of the subject learning outcomes. With precise prompts, ChatGPT seems to be able to work as an assistant to them in education, generating responses and deliverables in a quite structured and almost instant manner. In this paper, a ChatGPT-generated assessment script with its marking scheme and suggested solution of Probability and Engineering Statistics is explored based on Bloom’s Taxonomy. It is found that the ChatGPT-generated assessment script is partially complete and one of the multiple-choice questions is incorrect. The total score of the assessment script is not consistent with that of its marking scheme. Its suggested solution to one of the questions is missing. In addition, there are many application-oriented questions but few analysis-based questions and no evaluation-based question at all in the assessment script as per Bloom’s Taxonomy. Overall, ChatGPT-generated assessment scripts should further be reviewed and refined by educational practitioners to ascertain different levels of difficulty of questions which are in good alignment with the subject curriculum and the subject learning outcomes.