Page 267 - Kỷ yếu hội thảo khoa học lần thứ 12 - Công nghệ thông tin và Ứng dụng trong các lĩnh vực (CITA 2023)
P. 267
Thi-Thuy-Trinh Tran, Thi-Thanh-Minh Dang, Phuoc-Cuu-Long Le and Thi-Thu-Huong Dao 251
measurement model. Hence, the CFA results provided evidence of discriminant valid-
ity for all study constructs.
Table 2. The goodness of fit index
Fit index Obtained value Cut-off value
CMIN/df 2.402 <3
RMSEA 0.067 <=0.08
GFI 0.871 >=0.8
CFI 0.924 >=0.9
TLI 0.914 >=0.9
The results in table 2 showed that the model's suitability requirements could be ac-
cepted and declared fit. All five measurements showed a degree of good fit.
4.4 Hypotheses testing
Table 3. Hypothesis testing
Hypotheses Estimate S.E. C.R. P Result
H1: Attitude toward Social
Media Website Social .905 .081 11.110 0.000 Accepted
Media Experience
H2: Attitude toward the
University Brand Atti-
tude toward Social Media .456 .053 8.666 0.000 Accepted
Website
H3: eWOM usage behavior
Attitude toward the .945 .089 10.609 0.000 Accepted
University Brand
Table 3 and Figure 2 below showed that the hypotheses in the structural model are
statistically significant at the 0.05 level. Based on the results, it can be seen that the
variable Social Media Experience has a significant influence on Attitude towards the
Social Media Website, with impact coefficients 0.76. In addtion, the variable Attitude
towards the Social Media Website has a significant influence on Attitude towards the
University Brand (0.58), and Attitude towards the University Brand has an impact on
students' eWOM usage behavior (0.77).
ISBN: 978-604-80-8083-9 CITA 2023