This method marks a clear boundary between describing observable effects and guessing at causes beyond empirical evidence (hypotheses).
This method marks a clear boundary between describing observable effects and guessing at causes beyond empirical evidence (hypotheses).