Page 779 - ISC PROCEEDINGS 21.4

0.001), confirming that the correlation matrix is not an identity matrix. These results
demonstrate that the dataset is appropriate for factor analysis and supports the
extraction of latent constructs in the research model.
Factor extraction results (Total Variance Explained)s
After confirming data suitability through the KMO and Bartlett’s tests, exploratory
factor analysis was conducted using Principal Component Analysis (PCA) to determine the
number of factors representing the observed variables in the research model.
Table 3. Total Variance Explained
Extraction Sums of Rotation Sums of
Comp Initial Eigenvalues Squared Loadings Squared Loadings
onent % of Cumulative % of Cumulative % ofCumulative
Total Total Total
Variance % Variance % Variance %
1 6.805 42.532 42.532 6.805 42.532 42.532 3.364 21.027 21.027
2 2.068 12.926 55.459 2.068 12.926 55.459 2.968 18.548 39.574
3 1.538 9.613 65.072 1.538 9.613 65.072 2.844 17.772 57.346
4 1.200 7.499 72.571 1.200 7.499 72.571 2.436 15.224 72.571
5 .715 4.466 77.036
6 .665 4.158 81.194
7 .495 3.092 84.286
8 .436 2.723 87.010
9 .378 2.365 89.374
10 .356 2.227 91.602
11 .285 1.779 93.381
12 .274 1.711 95.091
13 .238 1.485 96.576
14 .206 1.291 97.866
15 .177 1.108 98.974
16 .164 1.026 100.000
Source: Author
Extraction Method: Principal Component Analysis.
Exploratory factor analysis using Principal Component Analysis (PCA) identified four
factors with eigenvalues greater than 1, satisfying Kaiser’s criterion. The first factor
explains 42.532% of the variance (eigenvalue = 6.805), followed by the second (12.926%),
third (9.613%), and fourth (7.499%). Together, these four factors account for 72.571% of
the total variance, exceeding the recommended threshold of 50% (Hair et al.), indicating
strong explanatory power of the model. After Varimax rotation, variance is more evenly
distributed across factors (21.027%, 18.548%, 17.772%, and 15.224%), suggesting a stable
and well-defined factor structure. Overall, the results confirm that the observed variables
are grouped into four factors consistent with the theoretical framework - digital
infrastructure, digital leadership, digital capability, and digital application - supporting the
convergent validity of the measurement model.
Rotated component matrix
After determining the number of factors based on the total variance explained,
Varimax rotation with Kaiser normalization was applied to clarify the factor structure and
improve interpretability. The rotation converged after six iterations, indicating a stable
factor solution.

778