Table 6.

Spearman correlations between sociodemographic characteristics of participants and their online health information–seeking behaviors.

Characteristics		Number of source typesa			Usage of LLMb-based chatbot for health informationsc					Number of topics searcheda					Cross-checking informationd					Following information and adviced
		ρ		P value	ρ		P value		ρ			P value		ρ			P value		ρ			P value
Age (years)		–0.1		.1	–0.16		.006		–0.11			.06		0.11			.07		0.12			.05
Education (years)		–0.02		.67	–0.06		.33		0.02			.68		–0.01			.82		–0.01			.86
Chronic health condition status		0.23		<.001	–0.01		.81		0.1			.09		0.02			.79		–0.05			.37
T-HCTe		–0.07		.23	–0.06		.32		–0.09			.13		0.02			.76		0.08			.21
eHEALSf		0.23		<.001	0.05		.38		0.11			.07		0.12			.04		0.14			.02
Artificial Intelligence Attitude Scale (AIAS)
	AIAS-4g	0.15	.01		0.36	<.001		0.18			.002		–0.13			.03		0.13			.04
	Medicine	0.14	.01		0.31	<.001		0.19			<.001		–0.06			.31		0.08			.16
Familiarity with ChatGPT		0.16		.01	0.3		<.001		0.24			<.001		0.01			.83		0.04			.55

^aSum of all reported by the participant.

^bLLM: large language model.

^cCoded as 0 if the participant reported not using large language model–based chatbots for health information and 1 if they have used them.

^dAverage across all sources reported by the participant. Values range from 0 to 1.

^eT-HCT: Trust in the Health Care Team.

^feHEALS: eHealth Literacy Scale.

^gAIAS-4: 4-item Artificial Intelligence Attitude Scale.