Page 705 - ISC PROCEEDINGS 21.4

2.4. Perceived usefulness and intention to use AI tools for learning
Perceived usefulness refers to the extent to which students believe that AI tools can
help them learn more effectively, save time, and improve academic outcomes. In
technology acceptance research, this construct is usually central because users are
unlikely to adopt a tool unless they recognize clear value for their tasks.
H2: Perceived usefulness (PU) positively affects intention to use AI tools for learning
(IU).
2.5. Social influence and intention to use AI tools for learning
Within the university environment, students’ decisions to use AI are shaped by the
classroom context, peer groups, and instructors. When students perceive that others
regard AI use as appropriate, useful, or worth encouraging, they are more likely to adopt
a favorable stance toward such tools.
H3: Social influence (SI) positively affects intention to use AI tools for learning (IU).
2.6. AI self-efficacy and intention to use AI tools for learning
AI self-efficacy captures students’ confidence in their ability to use AI tools
effectively for academic purposes. The more confident learners are in prompting,
interpreting outputs, adjusting content, and evaluating AI-generated information, the
more likely they are to use such tools proactively in learning.
H4: AI self-efficacy (AISE) positively affects intention to use AI tools for learning (IU).
2.7. Trust in AI and intention to use AI tools for learning
For generative AI in particular, trust is difficult to ignore. Students are more willing
to use AI tools for learning when they believe that the tool can provide reasonably
reliable support, generate meaningful content, and be used with an acceptable level of
safety in academic activities.
H5: Trust in AI (TR) positively affects intention to use AI tools for learning (IU).
2.8. Proposed research model
Based on the above hypotheses, the study tests the following regression model:
IU = β 0 + β 1PEOU + β 2PU + β 3SI + β 4AISE + β 5TR + ε
In this model, IU denotes intention to use AI tools for learning; PEOU denotes
perceived ease of use; PU denotes perceived usefulness; SI denotes social influence; AISE
denotes AI self-efficacy; and TR denotes trust in AI.
3. Research methodology
The study adopts a quantitative approach based on a structured questionnaire
survey. The dataset used for the analysis consists of 210 valid responses. With an initial
measurement model containing 30 observed variables across six constructs, this sample
size is acceptable for Cronbach’s Alpha, exploratory factor analysis, Pearson correlation,
and multiple regression analysis.
All constructs were measured using a five-point Likert scale ranging from 1 =
strongly disagree to 5 = strongly agree. The original measurement framework contained
six constructs: PEOU, PU, SI, AISE, TR, and IU. After reliability testing and factor-structure
review, five observed variables were removed from the model: SI3, AISE4, AISE5, TR3, and
TR4. The final analytical model therefore retained 25 observed variables. Each construct
was represented by the mean score of its retained items. The data were processed in
SPSS in four stages: (i) reliability testing using Cronbach’s Alpha, (ii) exploratory factor
analysis for the independent variables, (iii) Pearson correlation analysis, and (iv) multiple
linear regression to evaluate the influence of each factor on students’ intention to use AI
tools for learning.

704