search for: plainp

...groups (as is the case when both a generated from random draws from the normal distribution). I run an initial number of subjects in two groups (t1N) then, if p is < .25 on the t-test I add t2N more subjects to each group. Then I perform the t.test again. If the p was > .25 at time 1 I stop. Plainp is simply storing the p-values from t2 (if it was performed) or from t1 (if t2 was not performed). In the code I provide t1 starts at 16 since this is about when the problem becomes more frequent. Please note that it takes quite a long while to fail, and depending on what the true cause is it may n...