Table 4.

Pairwise comparison for nociception using “Tukey-Kramer” for a 95% confidence interval on a two-way ANOVA.

	Group 1	Group 2	Lower	Mean diff.	Upper	p-Value	Mean diff (%)
After learning	R	R + P	−2.4641	−0.9040	0.6560	0.4443	−22.90
	R	R + N	−0.6973	0.8627	2.4228	0.4863	21.85
	R	R + P + N	−2.2731	−0.7131	0.8470	0.6431	−18.07
	R + P	R + N	0.2068	1.7668	3.3268	0.0190	36.42
	R + P	R + P + N	−1.3691	0.1910	1.7510	0.9892	3.93
	R + N	R + P + N	−3.1358	−1.5758	−0.0158	0.0466	−51.08
During learning	R	R + P	−58.5381	−47.5636	−36.5891	0.0000	−49.86
	R	R + N	−8.0196	2.9549	13.9294	0.9003	3.10
	R	R + P + N	−36.5662	−25.5917	−14.6172	0.0000	−26.82
	R + P	R + N	39.5440	50.5185	61.4930	0.0000	35.34
	R + P	R + P + N	10.9973	21.9718	32.9463	0.0000	15.37
	R + N	R + P + N	−39.5212	−28.5466	−17.5721	0.0000	−30.88

Here, we analyze the average perceived nociception on the validation set of the best 4 hyperparameter sets for each condition, both after learning as well as the cumulative absolute value during learning. All hyperparameter sets were tested with 10 different randomly initialized networks. A positive percentage indicates the effect size in which Group 2 improves upon Group 1, whereas a negative percentage indicates the effect size in which Group 2 is worse than Group 1.