A Precrash Scenario Analysis Comparing Safety Performance across Autonomous Vehicle Driving Modes
Table 4
Results of one-way ANOVA and NPar test of precrash scenarios with different driving modes.
Precrash scenarios
Autonomous driving
Driving takeover
Conventional driving
ANOVA (mean ± standard deviation)
NPar test
N
%
N
%
N
%
Chi-square value
value
1
101
43.16
5
6.10
36
21.43
0.236 ± 0.186
69.491
≤0.001
2
31
13.25
13
15.85
20
11.90
0.137 ± 0.020
1.313
0.519
3
22
9.40
12
14.63
8
4.76
0.096 ± 0.049
11.851
0.003
4
12
5.13
8
9.76
17
10.12
0.083 ± 0.028
4.508
0.105
5
11
4.70
15
18.29
8
4.76
0.093 ± 0.078
31.508
≤0.001
6
9
3.85
8
9.76
7
4.17
0.059 ± 0.033
8.714
0.013
7
4
1.71
2
2.44
14
8.33
0.042 ± 0.036
15.200
0.001
8
8
3.42
4
4.88
6
3.57
0.040 ± 0.008
0.667
0.717
9
2
0.85
3
3.66
8
4.76
0.031 ± 0.020
6.091
0.048
10
6
2.56
2
2.44
4
2.38
0.025 ± 0.001
0.000
1.000
11
3
1.28
2
2.44
6
3.57
0.024 ± 0.011
2.235
0.327
12
2
0.85
2
2.44
7
4.17
0.025 ± 0.017
5.333
0.069
13
5
2.14
0
0.00
4
2.38
0.015 ± 0.013
0.091
0.763
14
2
0.85
0
0.00
5
2.98
0.013 ± 0.015
2.778
0.096
15
3
1.28
0
0.00
2
1.19
0.008 ± 0.007
0.000
1.000
16
0
0.00
3
3.66
2
1.19
0.016 ± 0.019
3.000
0.083
17
2
0.85
1
1.22
1
0.60
0.009 ± 0.003
1.000
0.607
18
3
1.28
0
0.00
1
0.60
0.006 ± 0.006
1.000
0.317
19
0
0.00
0
0.00
4
2.38
0.008 ± 0.014
—
—
20
2
0.85
0
0.00
1
0.60
0.005 ± 0.004
0.333
0.564
21
2
0.85
0
0.00
1
0.60
0.005 ± 0.004
0.333
0.564
22
1
0.43
1
1.22
1
0.60
0.007 ± 0.004
1.600
0.449
23
0
0.00
1
1.22
0
0.00
0.004 ± 0.007
—
—
24
3
1.28
0
0.00
5
2.98
0.014 ± 0.015
1.600
0.206
F = 4.338
Note. Reject the null hypothesis at the 5% level. The significance of the bold values is that the p value of the NPar test results is less than 5%, which indicates that the null hypothesis is rejected at the 5% level, and that the proportions of the different precrash scenario types show significant differences for the three driving modes overall.