Table 6.

Adjusted associations^a between mode of questionnaire completion (exposure) and 8 selected questions (outcomes; see Multimedia Appendix 2 on the Internet for full regression results).

Question type and question		Mode	Sociodemographic adjustedb		Sociodemographicb and propensity score adjusted
			Odds ratio or β coefficient	95% CI	Odds ratio or β coefficient	95% CI
Objective 1: social desirability
	Sensitive: More than 1 sexual partner in the last 12 months (n=1424)
		Telephone	1.00	—	1.00	—
		Web	3.76d	(1.86-7.59)	3.65d	(1.80-7.42)
	Stigma-related: afraid of people living with HIV or AIDS (n=2008)c
		Telephone	0.00	—	0.00	—
		Web	0.000	(−0.23-0.23)	0.019	(−0.21-0.25)
	Stigma-related: comfort with shopping at small grocery store owned by someone who has HIV or AIDS (n=1934)c
		Telephone	0.00	—	0.00	—
		Web	0.087	(−0.34-0.21)	0.069	(−0.05-0.19)
	Less-sensitive: Charitable giving in the past year (n=1996)
		Telephone	1.00	—	1.00	—
		Web	1.61d	(1.15-2.27)	1.63d	(1.15-2.29)
	Less-sensitive: perceived HIV knowledge (n=2022)c
		Telephone	0.00	—	0.00	—
		Web	−0.049	(−0.22-0.12)	−0.057	(−0.23-0.11)
Objective 2: Item Non-Response
	Sensitive: tested for HIV or AIDS (n=2027)
		Telephone	1.00	—	1.00	—
		Web	8.68d	(2.63-28.67)	8.04d	(2.46-26.31)
	Stigma-related: comfort with shopping at small grocery store owned by someone who has HIV or AIDS (n=2027)
		Telephone	0.00	—	0.00	—
		Web	2.99d	(1.41-6.32)	3.11d	(1.47-6.63)
	Less-sensitive: illness or disease that concerns you the most (n=2027)
		Telephone	1.00	—	1.00	—
		Web	3.01d	(1.68-5.38)	3.02d	(1.67-5.47)

^aSeparate logistic regression analyses were run for all questions except for continuous outcomes.

^bAdjusted for age, gender, education, household income, sexual minority status, and region.

^cLinear regression analysis.

^dIndicates statistical significance at the 95% confidence level.