A split sample experiment is conducted to test one of the most fundamental assumptions underpinning choice experiments in an environmental setting: whether stated values are absolute or relative. The test uses a natural experiment involving respondents at different home locations relative to potential new wetland sites. Respondents were presented with one of two ranges of distances. An absolute value interpretation requires that the functional relations with distance derived from these two ranges be consistent with each other. We reject a null hypothesis of no range bias, suggesting that respondents may perceive attribute levels in a relative rather than absolute sense.