An evaluation of statistical efficiency and bias trade-off involved with the use of follow-up questioning in the contingent valuation of environmental amenities