AI Misuse in Reading Assessments: A Case Study of Short-Story Reading in Higher Education

Abstract

This study examined higher education students’ experience and reading comprehension after reading short stories. Forty-one students were involved in this research and were chosen by using a convenience sampling method. The design of this research was a mixed-method research employing descriptive quantitative analysis and thematic interpretation. Two instruments were used, namely the reading comprehension test and the questionnaire, to evaluate students’ perception of short stories. The students’ reading comprehension mean score was 68.26 (categorized as unsatisfactory). Additionally, most students agreed that reading short stories broadened their vocabulary. Moreover, there was an unexpected finding in this research, which was not listed for research purposes, i.e., the misuse of Artificial Intelligence among students to answer reading questions. Furthermore, some students tried to use a translation tool from Google Lens; they copied the translation result on their answer sheet without any changes. The last group of students was caught cheating. They shared similar answers. Those academic violations influenced the reading comprehension means score due to a reduction in scores. In conclusion, reading short stories can be an alternative to teach reading since it opens opportunities to students to come across new vocabulary, which helps them to understand the story as well as grammar knowledge. To avoid the challenges found in this research, it is suggested that English teachers explain the pros and cons of the use of Artificial Intelligence in learning English, issue clear rules of AI usage in the classroom, and keep motivating the students to put more effort to learn English.