Table 2. Measurement model statistics – composite reliability, AVE, mean, loading range and discriminant validity of reflective constructs.

Constructs (final # of items)	CR^a	AVE^a	M (SD)	Loading range^a	Discriminant validity (Fornell-Larcker criterion) ^a
Constructs (final # of items)	CR^a	AVE^a	M (SD)	Loading range^a	AT	BI	PS	PV	RC	RE	SE	SN
AT (2)	0.806	0.675	4.21 (0.75)	0.82–0.83	0.822
BI (2)	0.880	0.710	4.15 (0.84)	0.82–0.86	0.586	0.843
PS (1)	-	-	4.38 (0.89)	-	0.267	0.294	1.000
PV (3)	0.797	0.567	3.82 (0.86)	0.72–0.80	0.416	0.503	0.324	0.753
RC (3)	0.896	0.742	2.47 (1.09)	0.85–0.87	–0.192	–0.274	–0.067	–0.225	0.861
RE (3)	0.791	0.558	4.27 (0.68)	0.73–0.76	0.586	0.391	0.212	0.405	–0.181	0.747
SE (2)	0.801	0.668	3.80 (0.93)	0.80–0.83	0.356	0.465	0.121	0.337	–0.297	0.322	0.817
SN (2)	0.835	0.717	3.86 (0.93)	0.84–0.85	0.392	0.540	0.152	0.466	–0.259	0.350	0.428	0.847

Acceptable values: Item loading >0.7, CR>0.7, AVE>0.5. To assess discriminant validity, the square root of each construct's AVE is compared to its correlations with other constructs.