Can I add additional readings on the same subject to the dataset, particularly when it is hard to find study subjects?
In most studies the sponsor should avoid multiple testing of the same subject because it may skew performance statistics and under-estimate standard deviations. When multiple testing is done, the sponsor should explain why it is necessary and choose methods of statistical analysis that allow adjustment for within-subject correlation. It is appropriate to conduct repeated testing of the same sample to evaluate test reproducibility – i.e., the ability of the test to yield the same or similar readings when expected.
