Skip to main content
. 2020 Aug 3;22(8):e19018. doi: 10.2196/19018

Table 2.

Source and content characteristics of smart assistant responses.

Variable Value (N=103), n (%)
Sourcea

Health information provider 36 (35.0)

Federal/city/state 25 (24.3)

Nonprofit/advocacy group 15 (14.6)

Physician 10 (9.7)

Other 9 (8.7)

News/media organization 8 (7.8)

Business 6 (5.8)

Journalist/press 2 (1.9)

Politician 2 (1.9)

Health care organization 1 (1.0)
Content

Format of the smart assistant responses


Video 30 (29.1)


General web page 24 (23.3)


Article 14 (13.6)


FAQb site 19 (18.4)


Other 6 (5.8)


Blog post 3 (2.9)


Location/map/directions 1 (1.0)

Topics mentioned in the smart assistant responsesa


Cancer 77 (74.8)


Vaccines and side effects 57 (55.3)


Sexual behaviors 26 (25.2)


Access to medical care 20 (19.4)


Direct impacts on loved ones 13 (12.6)


Conspiracy theories 6 (5.8)


Vaccines are ineffective 6 (5.8)


Mass media 6 (5.8)


Risk of disease is lower than risk of adverse vaccination events 2 (1.9)

Sentiment expressed toward vaccines in smart assistant responses


Positive 65 (63.1)


Neutral 18 (17.5)


Ambiguous 11 (10.7)


Negative 7 (6.8)

aTotals may exceed 100% owing to the availability of multiple response options.

bFAQ: frequently asked question.