Table 2.

Relative direct and indirect effects of the predictors on an individual’s intention to comply with a medical recommendation

Effects	ß	SE	p [95% CI]
Direct effects on the mediators
Constant	3.66	0.16	< 0.001
Physician → social presence	1.95	0.23	< 0.001
Physician plus AI → social presence	1.76	0.23	< 0.001
Disease severity → social presence	0.21	0.23	0.35
Physician × disease severity → social presence	− 0.32	0.32	0.32
Physician plus AI × disease severity → social presence	− 0.38	0.32	0.24
Constant	5.98	0.11	< 0.001
Physician → perceived innovativeness	− 0.41	0.15	0.002
Physician plus AI → perceived innovativeness	0.05	0.15	0.73
Disease severity → perceived innovativeness	0.07	0.15	0.66
Physician × disease severity → perceived innovativeness	− 0.27	0.21	0.20
Physician plus AI × disease severity → perceived innovativeness	0.01	0.21	0.98
Direct effects on the dependent variable
Constant	2.96	0.27	< 0.001
Physician → compliance intention	0.40	0.14	0.005
Physician plus AI → compliance intention	0.27	0.14	0.05
Social presence → compliance intention	0.11	0.03	< 0.001
Perceived innovativeness → compliance intention	0.37	0.04	< 0.001
Disease severity → compliance intention	0.15	0.13	0.24
Physician × disease severity → compliance intention	− 0.24	0.19	0.20
Physician plus AI × disease severity → compliance intention	− 0.40	0.19	0.03
Indirect effects via the mediators
Physician, total indirect effect	0.07	[− 0.12, 0.29]
Via social presence	0.22	[0.09, 0.39]
Via perceived innovativeness	− 0.15	[− 0.28, − 0.04]
Physician plus AI, total indirect	0.22	[0.03, 0.41]
Via social presence	0.20	[0.07, 0.36]
Via perceived innovativeness	0.02	[− 0.09, 0.12]

The experimental group that was assigned to the automated AI tool as the diagnostic method served as the reference group for comparison with dummy one (physician) and dummy two (physician plus AI)

ß = Unstandardized path coefficient, SE = Standard error, p = Significance, CI = Confidence interval