Table 7.
Analysis of mean support for use of artificial intelligence (AI) in specific scenarios, 95% CIs, and P values for testing against a mean of 3. A score <3 represents support, and a score of >3 represents oppositiona.
| Domain and scenario | Estimated mean (95% CI) | P valueb | Design effect | ||||
| General—support for the development of AI (B01c) | 2.35 (2.31-2.39) | <.001 | 1.83 | ||||
| Health | |||||||
|
|
Data sharing for quality care (C04) | 2.90 (2.83-2.98) | .01 | 1.65 | |||
|
|
Nonexplainable hospital algorithms (C03) | 3.25 (3.18-3.32) | <.001 | 1.57 | |||
|
|
Deskilling physicians (C05) | 3.39 (3.31-3.46) | <.001 | 1.62 | |||
| Welfare | |||||||
|
|
Targeted compliance checking (D03) | 2.98 (2.91-3.06) | .64 | 1.62 | |||
|
|
Automated parent support (contestability; D05) | 3.06 (2.99-3.13) | .10 | 1.60 | |||
|
|
Nonexplainable job services (D04) | 3.19 (3.12-3.26) | <.001 | 1.59 | |||
aMeans and CIs adjusted for weighting.
bP value for t test that the mean score was 3.0 using complex samples.
cCode in parentheses (eg, B01) indicates question number in instrument.