Trust and Quality in AI-Driven Tools: The Influence on Intention to Use ChatGPT Among University Students in Manila

Abstract

Universities are adopting conversational tools for coursework, yet evidence is mixed on what drives students’ intention to use them in higher education, especially in Metro Manila. Prior work rarely integrates credibility factors with belief-based predictors, so we test information quality and source trustworthiness alongside effort and performance expectancy to explain intention. Using a cross-sectional research design, primary data were collected through an online survey distributed via email and WhatsApp to students from Metro Manila. A total of 256 valid responses were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results reveal that source trustworthiness has a significant impact on both effort expectancy and performance expectancy, while effort expectancy has a strong predictive relationship with performance expectancy. In turn, performance expectancy is the most significant predictor of intention to use ChatGPT. However, the study found that information quality does not significantly influence performance expectancy, and effort expectancy does not directly influence intention to use. The study’s novelty is the integrated UTAUT–ELM model estimated on a Manila student sample, clarifying how credibility shapes usefulness beliefs and intention. Practically, instructors should emphasize transparent use, verification scaffolds, and course-aligned examples to raise perceived usefulness and credibility.

Ähnliche Arbeiten