Revisiting university students' intention to accept AI-Powered chatbot with an integration between TAM and SCT: a south Asian perspective
Authors:
- 28 March 2024
- journal article
- research article
- Published by Emerald Publishing in Journal of Applied Research in Higher Education
- Vol. 17 (2) , 594-608
Abstract
Purpose: This paper aims to explore students’ intention to use and actual use of the artificial intelligence (AI)-based chatbot such as ChatGPT or Google Bird in the field of higher education in an emerging economic context like Bangladesh.Design/methodology/approach: The present study uses convenience sampling techniques to collect data from the respondents. It applies partial least squares structural equation modeling (PLS-SEM) for analyzing a total of 413 responses to examine the study’s measurement and structural model.Findings: The results explore that perceived ease of use (PEOU) negatively affects intention to adopt AI-powered chatbots (IA), whereas university students’ perceived usefulness (PU) influences their IA positively but insignificantly. Furthermore, time-saving feature (TSF), academic self-efficacy (ASE) and electronic word-of-mouth (EWOM) have a positive and direct impact on their IA. The finding also reveals that students' IA positively and significantly affects their actual use of AI-based chatbot (AU). Precisely, out of the five constructs, the TSF has the strongest impact on students’ intentions to use chatbots.Practical implications: Students who are not aware of the chatbot usage benefits might ignore these AI-powered language models. On the other hand, developers of chatbots may not be conscious of the crucial drawbacks of their product as per the perceptions of their multiple users. However, the findings transmit a clear message about advantages to users and drawbacks to developers. Therefore, the results will enhance the chatbots’ functionality and usage.Originality/value: The findings of the study alert the teachers, students and policymakers of higher educational institutions to understand the positive outcomes and to accept AI-powered chatbots such as OpenAI’s ChatGPT. Outcomes also notify the AI-product developers to boost the chatbot’s quality in terms of timeliness, user-friendliness, accuracy and trustworthiness.Keywords
This publication has 54 references indexed in Scilit:
- Exploring the impact of augmented reality on student academic self-efficacy in higher educationComputers in Human Behavior, 2023
- Investigating student acceptance of an academic advising chatbot in higher education institutionsEducation and Information Technologies, 2023
- Role of Chat GPT in Public HealthAnnals of Biomedical Engineering, 2023
- Adaptation and validation of the Juvenile Victimization Questionnaire-R2 for a national study of child maltreatment in AustraliaChild Abuse & Neglect, 2023
- Understanding secondary students’ continuance intention to adopt AI-powered intelligent tutoring system for English learningEducation and Information Technologies, 2022
- Validation and translation of the Hungarian version of the Australian Pelvic Floor Questionnaire (APFQ-H)International Urogynecology Journal, 2022
- Quo vadis artificial intelligence?Discover Artificial Intelligence, 2022
- Video conferencing in the e-learning context: explaining learning outcome with the technology acceptance modelEducation and Information Technologies, 2022
- Preliminary Data Analysis: An Analysis Before the Analysis2021
- On the evaluation of structural equation modelsJournal of the Academy of Marketing Science, 1988