Table 3.

Between-participant comparisons for advice evaluations (Study 3). Table 3 presents the mean evaluations for overt ChatGPT-generated advice relative to Covert ChatGPT-generated advice in study 3. In both conditions, the advice was the same. We manipulated the source of the advice to participants. The columns represent the mean evaluations for our four dependent variables (“M”) and their corresponding standard deviations (“SD”). The right-hand side shaded portion of the table presents the results of between-subjects t-tests with the mean difference between the variables, with associated t-statistic, p-value, 95% confidence interval (“Lower CI” and “Upper CI”), and Cohen’s d effect sizes.

Variable	CONDITION				Mean diff.	t	p-value	Cohen’s d	Lower CI	Upper CI
	Overt ChatGPT		Covert ChatGPT
	M	SD	M	SD
Effectiveness	5.30	1.34	5.85	1.09	−0.55	5.21	.001	−0.45	0.34	0.76
Quality	5.41	1.19	5.74	1.04	−0.34	3.47	< .001	−0.30	0.15	0.53
Authenticity	5.02	1.46	5.78	1.10	−0.76	6.82	< .001	−0.59	0.54	0.99
Advice Uptake	5.32	1.29	5.72	1.14	−0.40	3.81	< .001	−0.33	0.20	0.61