This paper, grounded in the technology acceptance model (TAM) and the unified theory of acceptance and ue of technology (UTAUT), explores the impact of source credibility, assessed through expertise and trustworthiness, on the perceived usefulness of information and the ease of use in the context of AI-enhanced learning tools.
During 2024, 197 responses were collected from undergraduate and graduate business students at a Canadian university. The researchers developed a survey instrument with questions derived from prior research. The statistical tools used included SPSS, PLS-SEM, necessary condition analysis (NCA) and importance-performance map analysis (IPMA).
Four of five hypotheses were accepted, underscoring the importance of incorporating all constructs into the structural model of AI-enhanced learning tools. In the IPMA framework, findings suggest that perceived ease of use and expertise fell within the “No change” quadrant. In contrast, trustworthiness was located in the “Do better” quadrant of the IPMA map. Importantly, all exogenous constructs were identified as necessary, “must-have” conditions for the endogenous information usefulness construct, underscoring the importance of trustworthiness, particularly in AI-mediated learning contexts where credibility and evaluation of algorithm-generated content are critical.
The research recommends integrating the IPMA's analysis with the NCA using PLS-SEM. This paper highlights the importance of combining these methods with performance characteristics, especially when ranking activities that could enhance results in the target construct, such as the information usefulness in this study. While the combined use of PLS-SEM and NCA is a relatively recent approach, this study is the first to employ it with an IPMA framework in the context of AI-enhanced learning tools.
