Development of an AI-based Career Counseling Chatbot for University Students

Thesis Abstract

The increasing reliance on information and communication technology (ICT) in higher education necessitates innovative approaches to enhance career guidance services for university students, addressing persistent challenges related to accessibility, personalization, and timely support. This study aims to develop and evaluate an artificial intelligence (AI)-based career counseling chatbot tailored for university students, seeking to bridge gaps in traditional career guidance mechanisms. The specific objectives include designing a conversational AI system grounded in established career development theories, assessing its usability and effectiveness, and exploring its influence on students’ career decision-making confidence and career exploration behaviors. Employing a mixed-methods research design, the study integrates qualitative development processes with quantitative evaluation. The population comprises undergraduate students across three faculties—Arts, Sciences, and Business—at a mid-sized university, with a targeted sample size of 300 participants for quantitative analysis, selected via stratified random sampling to ensure representativeness across faculties and year levels. The qualitative component involves participatory design workshops with 20 students and 10 career guidance practitioners to inform chatbot development, ensuring alignment with user needs and pedagogical best practices. Data collection instruments include semi-structured interviews and focus group discussions for needs assessment and system design inputs, as well as standardized questionnaires for measuring user satisfaction, perceived usefulness, and career decision-making self-efficacy, adapted from established scales such as the Career Decision Self-Efficacy Scale (CDSE). The chatbot’s usability and engagement metrics are collected through system logs and user feedback surveys. Validity and reliability of the instruments are ensured through pilot testing, Cronbach’s alpha, and expert reviews. The analytical framework involves thematic analysis of qualitative data to identify key features for chatbot design, and quantitative analyses comprising descriptive statistics, t-tests, and multiple regression analyses. As a prelude to implementation, the Technology Acceptance Model (TAM) serves as the theoretical underpinning, complemented by Super’s Life-Span, Life-Space Theory, to guide the chatbot’s interaction design and assessment criteria. It is anticipated that the AI-based career counseling chatbot will demonstrate high usability scores and positive user perceptions, significantly improving students’ career exploration behaviors and confidence in decision-making, as evidenced by statistically significant increases in self-efficacy scores (p < 0.05). The findings are expected to reveal that personalized, accessible, and contextually relevant guidance via AI can effectively supplement existing career services, particularly in resource-constrained environments. This research contributes novel empirical evidence on the application of AI-driven solutions within career guidance in higher education, expanding understanding of their acceptability, efficacy, and scalability. It addresses a critical gap in technology-enhanced career support by presenting a replicable model that integrates pedagogical theories with emergent AI capabilities, thereby advancing both theoretical and practical knowledge in Guidance and Counselling. The study concludes with recommendations for integrating AI-based chatbots into institutional career frameworks, emphasizing the importance of iterative design, user-centered approaches, and data privacy considerations, alongside suggestions for further research on long-term impacts and cross-cultural applicability.