Page 542 - ISC PROCEEDINGS 21.4

authors further assessed convergent validity using the Average Variance Extracted (AVE)
for the constructs FL, SA, PP, MA, BU, and SI. All AVE values exceeded the threshold of 0.5,
thereby confirming adequate convergent validity.
4.2.2. Discriminant validity
We confirmed discriminant validity through the Fornell-Larcker criterion along with
the heterotrait-monotrait (HTMT) ratio. Each construct's square root of AVE was greater
than its correlations to other constructs, thus fulfilling the Fornell-Larcker standard
(Fornell & Larcker, 1981). Moreover, every HTMT value stayed under the strict 0.85 cutoff,
offering strong support for discriminant validity (Henseler et al., 2015).
Table 5. Discriminant validity (Fornell-Larcker criterion)
FL SA PP MA BU SI
FL 0.780
SA -0.583 0.823

PP 0.577 -0.667 0.824
MA -0.657 0.557 -0.572 0.758
BU -0.482 0.683 -0.582 0.509 0.786
SI -0.641 0.662 -0.637 0.576 0.608 0.784
Source: Author
Table 6. Discriminant validity (Heterotrait-Monotrait correlation ratio)
FL SA PP MA BU SI

FL
SA 0.669
PP 0.708 0.811
MA 0.762 0.640 0.705
BU 0.549 0.771 0.707 0.584

SI 0.731 0.752 0.778 0.670 0.690
Source: Author
4.3. Structural model assessment
4.3.1. Multicollinearity
Multicollinearity was assessed using the Variance Inflation Factor (VIF).
Table 7. Multicollinearity statistics (VIF)
FL FLxBU SA PP MA BU SI
FL 1.748
FLxBU 1.285
SA 2.602

PP 2.087
MA 1.000
BU 1.000 1.000 1.991
SI
Source: Author
All VIF values were below the threshold of 3.0, indicating that collinearity does not
pose a concern in the structural model (Hair et al., 2011).

541