Table 3.
Construct reliability and validity
| Construct | Items | loadings | CR | AVE | Cronbach’s α | VIF |
|---|---|---|---|---|---|---|
|
Perceived convenience (PCON) |
PCON1 | 0.81 | 0.89 | 0.66 | 0.89 | 1.91 |
| PCON2 | 0.71 | |||||
| PCON3 | 0.84 | |||||
| PCON4 | 0.88 | |||||
| Perceived complementarity (PCOM) | PCOM1 | 0.81 | 0.82 | 0.60 | 0.81 | 1.89 |
| PCOM2 | 0.78 | |||||
| PCOM3 | 0.74 | |||||
|
Information overload (IO) |
IO1 | 0.83 | 0.89 | 0.68 | 0.89 | 1.96 |
| IO2 | 0.86 | |||||
| IO3 | 0.84 | |||||
| IO4 | 0.75 | |||||
|
System feature overload (SYO) |
SYO1 | 0.73 | 0.88 | 0.66 | 0.88 | 2.00 |
| SYO2 | 0.74 | |||||
| SYO3 | 0.89 | |||||
| SYO4 | 0.87 | |||||
|
Flow experience (FE) |
FE1 | 0.81 | 0.90 | 0.68 | 0.89 | 2.57 |
| FE2 | 0.86 | |||||
| FE3 | 0.80 | |||||
| FE4 | 0.84 | |||||
|
Self efficacy (SEEF) |
SEEF1 | 0.74 | 0.88 | 0.66 | 0.88 | 1.52 |
| SEEF2 | 0.81 | |||||
| SEEF3 | 0.83 | |||||
| SEEF4 | 0.86 | |||||
|
Cognitive engagement (CE) |
CE1 | 0.81 | 0.82 | 0.60 | 0.82 | 1.68 |
| CE2 | 0.81 | |||||
| CE3 | 0.69 | |||||
|
Emotional engagement (EE) |
EE1 | 0.89 | 0.94 | 0.80 | 0.94 | 2.35 |
| EE2 | 0.92 | |||||
| EE3 | 0.89 | |||||
| EE4 | 0.88 | |||||
|
Behavioural engagement (BE) |
BE1 | 0.72 | 0.79 | 0.55 | 0.79 | 1.91 |
| BE2 | 0.81 | |||||
| BE3 | 0.70 | |||||
|
Learning effectiveness (LE) |
LE1 | 0.89 | 0.90 | 0.70 | 0.90 | / |
| LE2 | 0.85 | |||||
| LE3 | 0.78 | |||||
| LE4 | 0.84 |
Furthermore, the correlation coefficients of all structures were compared with the square root of the AVE value. The results demonstrated that the correlation coefficients of each structure were less than the square root of the AVE value, indicating that these structures had good discrimination validity (Tenko Raykov, 2010). Table 4 presents the results of this study