When scientist do their experimental test more than once so they can reduce the effects of chance errors what is this called in one word?

replication