Table 4.

Regression models comparing patients with SCZ and HC.

Model specification	Fixed effects	b	SE	t	p-value
Sensitivity
dprime ~ Communicative phenomenon + Group + Age + Gender + Level of intelligence + (1 | ID)	Sincere	3.15	0.14	22.75	<0.001
	Deceit	2.91	0.14	21.01	<0.001
	Irony	2.93	0.14	21.18	<0.001
	Level of intelligence	0.27	0.07	3.71	<0.001
	Age	−0.14	0.07	−1.97	0.055
	Gender	−0.29	0.16	1.82	0.076
	Sincere: Group = SCZ	−0.41	0.19	−2.17	<0.05
	Deceit: Group = SCZ	−0.86	0.19	−4.51	<0.001
	Irony: Group = SCZ	−0.67	0.19	−3.54	<0.001
Response bias
Beta values ~ Communicative phenomenon + Group + Age + Gender + Level of intelligence + (1 | ID)	Sincere	0.78	0.11	6.82	<0.001
	Deceit	0.83	0.12	6.93	<0.001
	Irony	0.89	0.12	7.04	<0.001
	Level of intelligence	0.02	0.06	0.29	0.76
	Age	0.01	0.01	0.09	0.93
	Gender	−0.11	0.12	−0.94	0.35
	Sincere: Group = SCZ	−0.05	0.14	−0.35	0.72
	Deceit: Group = SCZ	0.40	0.19	2.07	<0.05
	Irony: Group = SCZ	−0.19	0.15	−1.29	0.20

b beta regression coefficient, SE standard error, t t-value. The response bias model was fitted using a Gamma distribution with an inverse link, thus positive beta regression coefficients (e.g., 0.40 for deceit in group of patients with SCZ) indicate lower beta values.