. Author manuscript; available in PMC: 2024 Sep 18.

Published in final edited form as: Cognit Ther Res. 2023 Jul 4;47(5):841–850. doi: 10.1007/s10608-023-10408-2

Table 3.

Multilevel modeling of AS, EA, and anxiety

	MASC-C			MASC-P			CGI-S
Model	Separate Models	Joint Model		Separate Models	Joint Model		Separate Models	Joint Model
	unstandardized	unstandardized	standardized	unstandardized	unstandardized	standardized	unstandardized	unstandardized	standardized
Time	−16.02 (2.05) ^***			−13.07 (1.87) ^***			−.89 (.11) ^***
AFQ-Y	.68 (.11) ^***	.35 (.15) ^*	5.09 (2.2) ^*	.16 (.10)	−.07 (0.15)	−1.10 (2.17)	.01 (.00) ^*	.01 (.01)	.11 (.09)
CASI	1.31 (.19) ^***	.85 (.27) ^**	6.80 (2.2) ^**	.50 (.19) ^*	.60 (.27) ^*	4.80 (2.17) ^*	.02 (.00) ^*	.01 (.01)	.06 (.09)
Time * AFQ-Y	−.32 (.20)			−.12 (.18)			.00 (.00)
Time * CASI	.13 (.36)			.21 (.33)			.00 (.00)

Note. Statistics above list b with SE in parathesis

= p < .05

^**

= p < .01

^***

= p < .001

Separate models tested each predictor separately; the AFQ and CASI models controlled for time. In the joint model, the AFQ, CASI, and time were all regressed on each outcome variable.