Table 1. Summary table of yes distributions in Experiments 1–4.

Per cent explained variance and statistical indices of goodness of fit of glm-based vs. observed yes distributions in Experiments 1–4.

			Best glm fitting parameters
	Per cent explained variance	Goodness of fit	β1	β2
Experiment 1 (uncomfortable ⇒ comfortable)	86	F(1, 78) = 504.0, p < 0.001	0.002 ± 0.035, t = −0.059, p = 0.95	1.00 ± 0.045, t vs. 1 = 0.02, p = 0.98
Experiment 2 (comfortable ⇒ uncomfortable)	88	F(1.78) = 595.7, p < 0.001	0.005 ± 0.031, t = 0.17, p = 0.86	0.99 ± 0.040, t vs. 1 = 0.18, p = 0.85
Experiment 3 (uncomfortable ⇒ comfortable)	86	F(1, 78) = 551.9, p < 0.001	0.033 ± 0.030, t = −1.07, p = 0.28	1.05 ± 0.044, t vs. 1 = 0.77, p = 0.44
Experiment 4 (comfortable ⇒ uncomfortable)	73	F(1, 70) = 185.2, p < 0.001	0.001 ± 0.05, t = 0.17, p = 0.87	0.98 ± 0.070, t vs. 1 = 0.15, p = 0.87