Alpha-error inflation due to optional stopping in data collection for an undirected t-test for differences between means of two-groups given no effect (i.e., H0 is true).
Initial number of participants for each group
5
50
0
0
20
5
10
15
20
25
30
35
40
45
50
Additional number of participants for each group before testing
0
20
0
0
5
0
2
4
6
8
10
12
14
16
18
20
Maximum number of participants for each group
10
100
0
0
50
10
20
30
40
50
60
70
80
90
100
Repetitions Monte-Carlo simulation
100
1,000
0
0
500
100
200
300
400
500
600
700
800
900
1,000
After changing parameters push *Run* to start simulation.
Run
Boxplot of p-values < .05
Data simulations