Transparency Enhances Positive Perceptions of Social Artificial Intelligence
Table 2
Participant background information.
Overall sample
Human control
Nontransparent intelligent
Nontransparent machine
Transparent intelligent
Transparent machine
ANOVA/Chi2
Age
36.85 (10.51)
37.42 (10.58)
36.26 (10.61)
36.79 (10.21)
36.36 (10.6)
37.46 (10.65)
,
Prior AI knowledge
10.32 (6.05)
10.13 (5.96)
10.78 (6.19)
10.32 (5.89)
10.15 (5.83)
10.25 (6.41)
,
Gender
,
Female
35.40%
30.94%
37.16%
38.17%
34.05%
36.87%
Male
64.20%
69.06%
62.30%
61.29%
65.41%
63.13%
Non-binary
0.30%
0.00%
0.55%
0.54%
0.54%
0.00%
Race/ethnicity
,
American Indian/native
0.66%
0.00%
0.00%
1.61%
0.54%
1.12%
Asian or Pacific Islander
4.49%
5.52%
3.83%
3.23%
6.49%
3.35%
Black or African American
7.77%
11.05%
8.20%
5.91%
5.41%
8.38%
Hispanic or Latino
2.30%
3.31%
2.19%
2.15%
2.16%
1.68%
Multiracial
1.86%
1.66%
2.19%
2.69%
0.54%
2.23%
White
82.82%
78.45%
83.06%
84.41%
84.86%
83.24%
Other
0.11%
0.00%
0.55%
0.00%
0.00%
0.00%
Education level
,
Graduate degree
17.07%
18.78%
16.39%
20.43%
14.05%
15.64%
4-year college degree
58.21%
57.46%
61.20%
53.76%
61.62%
56.98%
Some college or vocational
15.32%
15.47%
12.57%
12.90%
17.30%
18.44%
High school graduate
9.08%
8.29%
9.84%
12.37%
7.03%
7.82%
No high school degree
0.33%
0.00%
0.00%
0.54%
0.00%
1.12%
Income
,
$150,000 or more
1.75%
0.55%
2.73%
2.73%
1.08%
1.68%
$100,000 to $149,999
11.27%
13.81%
10.38%
10.38%
10.81%
12.85%
$50,000 to $99,999
45.07%
43.65%
42.62%
42.62%
48.11%
45.81%
$25,000 to $49,999
28.23%
26.52%
30.60%
30.60%
27.57%
26.82%
Less than $25,000
13.68%
15.47%
13.66%
13.66%
12.43%
12.85%
AI-related occupations
,
No
41.36%
38.67%
40.44%
40.86%
45.95%
40.78%
Yes
58.64%
61.33%
59.56%
59.14%
54.05%
59.22%
Chatbot use frequency
,
Never
1.97%
0.55%
3.27%
2.69%
0.00%
3.35%
Less than monthly
13.12%
16.02%
13.66%
12.90%
9.19%
13.97%
Monthly
35.23%
34.81%
33.87%
34.95%
37.30%
34.80%
More than weekly
49.67%
48.62%
49.18%
49.46%
53.51%
47.49%
Observations ()
914
181
183
186
185
179
Note. For numeric variables such as age and prior AI knowledge, the standard deviation is in parentheses. Prior AI knowledge was measured based on participants’ self-reported familiarity with seven AI-related terminologies on a scale from 0 (“I’ve never heard of this term”) to 3 (“I have a good understanding of how it works”), with a maximum score of 21.
