Augmenting short Cheap Talk scripts with a repeated Opt-Out Reminder in Choice Experiment surveys

Hypothetical bias remains an essential problem when valuing non-market goods with stated preference methods. Originally developed for contingent valuation studies, cheap talk has been found to effectively reduce hypothetical bias in some applications, though empirical results are ambiguous. We discuss two potential reasons why cheap talk might be ineffective, especially in a choice experiment setting. Particularly, cheap talk may fail to account for the repeated choice nature of choice experiments, and the applied wording in cheap talk scripts is typically very indirect in addressing hypothetical bias and only implicitly instructing respondents to avoid it. In the light of these potential shortcomings, we suggest adapting cheap talk to the choice experiment format by augmenting it with a so-called Opt-Out Reminder. Prior to each single choice set, the opt-out reminder explicitly instructs respondents to choose the opt-out if they find the experimentally designed alternatives too expensive. In a choice experiment survey we find support of the arguments regarding the potential shortcomings of cheap talk. Rather than merely adopting the cheap talk practice directly from contingent valuation, it should be adapted to fit the choice experiment format. Augmenting cheap talk with a dynamic opt-out reminder proves to be an effective and promising improvement in the continued effort to remedy hypothetical bias in choice experiment surveys.