. 2022 Mar 2;79(4):313–322. doi: 10.1001/jamapsychiatry.2022.0051

Table. Results From the Multivariate Analysis of Variance^a.

Analysis method	Main association	Approximate F	df		P value
Analysis method	Main association	Approximate F	Numerator	Denominator	P value
VBA
1^b	Group	971.30	16	15 273 788	<.001
	Task	60 255.67	64	61 095 164	<.001
	Study	14 795.04	416	244 380 848	<.001
2^c	Group	10 590.50	16	15 273 788	<.001
	Task	47 039.51	64	61 095 164	<.001
	Study	11 683.04	416	244 380 848	<.001
MAP
1^b	Group	672.91	16	15 273 788	<.001
	Task	20 178.88	64	61 095 164	<.001
	Study	3636.98	416	244 380 848	<.001
2^c	Group	1707.41	16	15 273 788	<.001
	Task	20 696.29	64	61 095 164	<.001
	Study	6560.50	416	244 380 848	<.001

Abbreviations: MAP, maximum a posteriori; VBA, variational bayesian analysis.

^{^a}

Full univariate results for all parameters can be found in the eResults in the Supplement.

^{^b}

Parameters were estimated using a hierarchical bayesian approach with only a single prior over groups.

^{^c}

Parameters were estimated using a different prior for each group.

Table. Results From the Multivariate Analysis of Variancea.