Page 648 - ISC PROCEEDINGS 21.4

To ensure ethical compliance:
Participation was anonymous.
No personal identifying information was collected.
Respondents provided informed consent before completing the survey. Data were
used strictly for academic research purposes.
3.5. Data analysis techniques
Data analysis was conducted using Structural Equation Modeling (SEM) to test the
proposed hypotheses and examine the relationships among variables.
The analysis followed a two-step approach:
3.6. Measurement model assessment
Model Assessment Criteria / Indicators Threshold / Method
Measurement Model
Assessment Cronbach’s Alpha ≥ 0.70
Composite Reliability (CR) ≥ 0.70
Average Variance Extracted (AVE) ≥ 0.50
Factor Loadings ≥ 0.70
Fornell–Larcker Criterion
Discriminant Validity
and HTMT Ratio
Structural Model Path Coefficients β values
Assessment
Significance Testing t-values and p-values
Coefficient of Determination R²
Effect Size f²
Predictive Relevance Q²
Statistical Procedure Bootstrapping 5,000 resamples
Software Used SEM Software SmartPLS or AMOS
3.7. Ethical considerations
The study followed ethical research standards in accordance with university
guidelines. Participants were informed about the voluntary nature of participation and
their right to withdraw at any time. Data confidentiality and anonymity were strictly
maintained.
4. Results
4.1. Respondent profile
A total of 342 valid responses were collected from university students. After data
screening, no significant missing values or outliers were identified. The sample consisted
of 52% female and 48% male students. Approximately 63% were undergraduate students
and 37% were postgraduate students. More than 70% reported using generative AI tools
(e.g., ChatGPT) at least once per week.
4.2. Measurement model assessment
The measurement model was evaluated using reliability and validity tests, including
factor loadings, Cronbach’s alpha, composite reliability (CR), average variance extracted
(AVE), and discriminant validity (HTMT ratio).
4.3. Reliability and convergent validity
All factor loadings exceeded the recommended threshold of 0.70. Cronbach’s alpha
and composite reliability values were above 0.70, and AVE values exceeded 0.50,
confirming convergent validity.

647