Table 3.

Model comparisons for summed base image and category oddball harmonics

	P(M)	P(M|data)	Log(BFM)	Log(BF10)	error %
Base image response models
Null model (subject + random slopes)	0.2	0.319	0.628	0.000
Group	0.2	0.342	0.733	0.070	2.771
Stim. Type	0.2	0.131	− 0.506	− 0.890	7.020
Stim. Type + Group	0.2	0.117	− 0.637	− 1.005	7.708
Stim. Type + Group + Stim. Type ✻ Group	0.2	0.091	− 0.916	− 1.225	1.832
Category oddball response models
Null model (subject + random slopes)	0.2	2.142 × 10–4	− 7.063	0.000
Stim. Type	0.2	0.579	1.705	7.903	1.029
Stim. Type + Group	0.2	0.269	0.387	7.137	1.273
Stim. Type + Group + Stim. Type ✻ Group	0.2	0.152	− 0.335	6.564	5.193
Group	0.2	9.274 × 10–5	− 7.899	− 0.836	1.497

Exhaustive comparisons of possible models in the Category Selectivity experiment, conducted separately for base image (top) and category oddball (bottom) responses, containing every possible combination of factors (including main/interaction terms), against the null model including only subject and random slope terms for all repeated measures factors