Starting Point Bias in Dichotomous Choice Valuation with Follow-Up Questioning

Follow-up questions are frequently used to improve the efficiency of dichotomous choice contingent valuation questionnaires. However, a number of authors (e.g., [12, 16]) have noted a significant difference between the WTP distributions implied by initial and follow-up question responses. This paper investigates starting point bias as one explanation for the phenomenon. We develop a model of starting point bias in the context of dichotomous choice questioning with follow-up, using a Monte Carlo simulation to illustrate the potential bias imparted to WTP estimates. An application is provided based on data from a CV study in northcentral Iowa.