The closeness of agreement among repeated measurements of a variable made under the same operating conditions over a period of time, or by different people.
The closeness of agreement among scientific results more generally, at the level of whole experiments, either nearly identical or similar.