| Sample size for t-test (per group) |
Percent of missing data |
Percent of 1,000 rejected Ho hypothesis
|
| Gold standard with no missing data |
DBI |
Percent from gold standard |
Mean based imputation |
Percent from gold standard |
| 50 |
10 |
39.0 |
39.3 |
0.3 |
42.4 |
3.4 |
|
20 |
36.3 |
40.5 |
4.2 |
46.0 |
9.7 |
|
30 |
36.8 |
38.0 |
1.2 |
48.4 |
11.6 |
| 100 |
10 |
64.3 |
63.0 |
–1.3 |
67.5 |
3.2 |
|
20 |
66.2 |
65.3 |
–0.9 |
72.6 |
6.4 |
|
30 |
67.2 |
64.5 |
–2.7 |
74.9 |
7.7 |
| 150 |
10 |
83.2 |
80.5 |
–2.7 |
84.1 |
0.9 |
|
20 |
84.2 |
80.4 |
–3.8 |
85.6 |
1.4 |
|
30 |
80.4 |
79.6 |
–0.8 |
85.9 |
5.5 |
| 200 |
10 |
92.2 |
89.2 |
–3.0 |
92.4 |
0.2 |
|
20 |
91.4 |
89.4 |
–2.0 |
93.4 |
2.0 |
|
30 |
92.2 |
90.7 |
–1.7 |
95.0 |
2.8 |