Reply to Schenkerman's avoiding rank reversal in AHP decision support models