Factors Influencing Engineering Students’ Perceptions on the Use of ChatGPT

Abstract The language model known as Chat Generative Pre-Trained Transformer (ChatGPT) was developed by Open Artificial Intelligence engineers. It's a kind of AI system that can produce text responses to a variety of questions and prompts that seem human. ChatGPT provides a number of benefits, such as round-the-clock assistance, prompt question answering, research-related information discovery, coding program writing, etc. Notwithstanding these benefits, ChatGPT's limited contextual knowledge of a given subject may result in inaccurate or irrelevant responses. Additionally, the feedback may be unfair or erroneous due to bias in the data used to train the program. Sadly, ChatGPT has the potential to pose security risks, which could result in data breaches and the leakage of private student information. This research project aims at understanding the factors influencing engineering students' perceptions on the use of ChatGPT. This topic is relevant, timely, and important as ChatGPT as created sufficient stir in education. By exploring factors influencing students' experiences and perspectives, we aim to shed light on different aspects of the usage of ChatGPT and glean critical insights. This research study answers the following research question, 'What factors influence the engineering students' perceptions on the use of ChatGPT?' A survey instrument was designed which included five dimensions: learning tool (10 items), trustworthiness (5 items), ethical considerations (5 items), ease of access (6 items), and concerns with ChatGPT (6 items). Additionally, the survey instrument also included demographic questions such as gender identity, race/ethnicity, current engineering academic department, and class standing. Four factors were identified by exploratory factor analysis (EFA): learning tool, trustworthiness, ease of access, and concerns with ChatGPT. Following the EFA, it was suggested that the dimension "ethical considerations" be eliminated. The range of Cronbach's alpha was 0.62 to 0.82, indicating a high degree of internal consistency reliability among the items. The statistical analysis reveals that males reported higher self-efficacy in using ChatGPT as a learning tool in comparison with other gender identities. Furthermore, Freshmen engineering students tend to have high perceptions on using ChatGPT as a learning tool, while junior engineering students have the lowest. Finally, freshmen engineering students tend to have high perceptions on ease of accessing ChatGPT, while sophomore engineering students have the lowest.