Table 1 Details of Bayesian statistical tests.

Reference	Condition	Mean (Δ%)	SD (Δ%)	DF	H1	B₁₀
Same color	Small family	−45.41	10.45	29.25	~ HN(34.58) ≤ 0	494.92
	Distinct colors	−29.52	10.33	31.01	~ HN(34.58) ≤ 0	17..33
	Frequent outlier	−10.79	10.50	29.16	~ HN(34.58) ≤ 0	0.80
	Similar colors	−1.34	7.59	37.97	~ HN(34.58) ≤ 0	0.25
	Concurrent	5.55	8.97	34.80	~ HN(34.58) ≤ 0	0.17
					~ HN(4.24) ≥ 0	1.13
	Nonlinear	9.89	8.17	37.05	~ HN(34.58) ≤ 0	0.11
					~ HN(4.24) ≥ 0	1.48
Distinct colors	Similar colors	28.18	10.41	31.54	~ HN(15.71) ≥ 0	10.80
	Added noise	28.00	11.18	35.62	~ HN(15.71) ≥ 0	7.76
	One-by-one	15.46	12.25	37.91	~ HN(15.71) ≥ 0	1.70
	Nonlinear+	12.00	12.19	37.87	~ HN(15.71) ≥ 0	1.30
	Speeded response	8.83	12.36	37.97	~ HN(15.71) ≥ 0	1.03
	Concurrent+	7.60	12.99	37.83	~ HN(15.71) ≥ 0	0.95
Same/distinct	Nonlinear/nonlinear+	10.95	7.25	78.00	~ HN(11.51) ≥ 0	2.16

Each row is one comparison. The data (likelihood) model was a Student’s t distribution with mean equal to the change in categorical encoding strength from reference to target condition, standard deviation estimated assuming unequal variances, and Welch-Satterthwaite degrees of freedom. The alternative hypothesis model was a half-normal distribution (HN) with standard deviation determined by the results of the Reference condition (see Methods). In all cases, the null hypothesis model was a point distribution at zero.

Quick links

Search