2004 Volume 21 Pages 265-275
When a respondent is faced with a two-scenario willingness-to-pay question, inconsistencies in answer are deemed to happen because of failure to discriminate between scenarios and elicitation format. This study investigates response biases by single and joint estimation of scenarios to account for correlation between responses. The models are applied to the valuation of the environmental impact of a road project with two alternative alignments on a cultural heritage site. Results show that if not addressed properly, these biases can cause misestimating contingent values.