In medical statistics, what is usually meant by "internal precision"? I have come across the term "internal precision" in some books and papers and was wondering what this usually would refer to. Is there a standard definition for this term? Thanks!
 A: In medical contexts (perhaps also in other contexts), accuracy and precision are often used to describe the concepts of validity and reliability:

Accuracy in this sense refers to the average of the "hits" being centered around the intended target (bulls-eye) and precision refers to the clustering of "hits" around the target. The analogy is that validity refers to if you're measuring the thing you intend to measure (how close to the target you get), and reliability refers to how consistent your measures are, regardless of the validity.
In some contexts reliability is divided in internal and external reliability. Internal reliability regards the consistency of the results within an individual or a study, and external reliability regards the consistency of the result between individuals or studies. Two examples:
If a self-assessment test is administered  the external reliability describes the tendency of the individual to give the same results on several test occasions, and the internal reliability is the tendency to give similar results within the same test. This means that items that reflect the same aspects should give approximately the same results.
In another context, internal reliability may refer to the consistency of the results in the study. This means that if somebody else would analyze your data, they would reach the same conclusion as you did. External reliability in this context refers to the consistency between studies, so that another research group that repeats your study gets similar results.
I have not come across the term internal precision, but I would guess that it refers to internal reliability. Does this make sense?
