Table 2. CVPAT results for CUSA and CUSL...

Table 2.

CVPAT results for CUSA and CUSL

Data set	PLS-SEM loss	Equal weight loss	Difference	p-value
Corprep22	0.629	0.630	−0.001	0.493
Corprep23	0.555	0.559	−0.004	0.003

Data set	PLS-SEM loss	Equal weight loss	Difference	p-value
Corprep22	0.629	0.630	−0.001	0.493
Corprep23	0.555	0.559	−0.004	0.003

Notes:

Loss = average loss. The calculation follows the $C V P A T_{c o n s t r u c t}^{c o m p a r e}$ measure from Sharma et al. (2023), but instead of testing the out-of-sample predictive ability between two PLS-SEM, the out-of-sample predictive ability is tested between a PLS-SEM and an equal weight model. p-values are based on 5,000 bootstrap samples

Source: Authors’ own work

[ViewLarge]