Well not everyone starts experiment anew. Many also reuse accumulated datasets. For human data even more so.