Table 1.

Results of individual model analysis^a

		Pearson correlation between RDMs				Noise ceiling (lower bound, upper bound)
		Mean r	SE	Z	p < 0.05 (FDR-corrected)
rFFA						0.135, 0.262
	Perceived Similarity	0.109	0.023	3.689	Yes
	Social Traits (All)	0.104	0.031	2.710	Yes
	OpenFace	0.101	0.023	3.461	Yes
	Attractiveness	0.090	0.033	2.687	Yes
	Gender	0.086	0.021	3.302	Yes
	Valence	0.060	0.023	2.391	Yes
	Dominance	0.058	0.030	1.640	No
	Gabor-Jet	0.052	0.049	0.956	No
	Trustworthiness	0.040	0.029	1.594	No
	Pixel-Faces	0.035	0.044	0.865	No
	Pixel-Frames	0.005	0.027	0.159	No
	GIST-Faces	−0.006	0.040	0.114	No
	Pixel-Frames	−0.018	0.041	−0.478	No
rOFA						0.337, 0.408
	Pixel-Faces	0.221	0.031	4.357	Yes
	Gabor-Jet	0.204	0.037	3.968	Yes
	Pixel-Frames	0.107	0.031	3.016	Yes
	GIST-Faces	0.104	0.043	2.216	Yes
	Attractiveness	0.092	0.029	2.843	Yes
	Social Traits (All)	0.083	0.031	1.979	No
	Gender	0.074	0.021	2.757	Yes
	OpenFace	0.067	0.020	2.952	Yes
	Dominance	0.055	0.031	1.546	No
	Perceived Similarity	0.039	0.026	1.416	No
	GIST-Frames	0.025	0.034	0.746	No
	Trustworthiness	0.011	0.025	0.400	No
	Valence	−0.016	0.031	−0.573	No
rpSTS						0.126, 0.252
	GIST-Frames	0.075	0.047	1.800	No
	Dominance	0.052	0.027	1.800	No
	OpenFace	0.040	0.020	2.129	No
	Social Traits (All)	0.032	0.026	1.018	No
	Pixel-Frames	0.022	0.030	0.956	No
	Gender	0.020	0.017	0.956	No
	Trustworthiness	0.017	0.032	0.524	No
	Attractiveness	0.005	0.024	0.134	No
	Valence	0.002	0.031	0.051	No
	Pixel-Faces	−0.003	0.035	−0.113	No
	Perceived Similarity	−0.008	0.026	−0.072	No
	Gabor-Jet	−0.045	0.040	−1.100	No
	GIST-Faces	−0.048	0.036	−1.368	No

^a Values correspond to the results presented in Figure 3A. For each ROI, we show the mean correlations between brain RDMs with each model, SE, Z statistics from two-sided one-sample Wilcoxon signed-rank tests, and whether correlations were significantly >0. We also show the estimated lower and upper bounds of the noise ceiling for each ROI. Models are ordered by effect size.