Information Processing in Stated Preference Surveys \Newline a Case Study on Urban Gardens

Stated preference surveys must provide information on the good to be valued, and respondents must process and recall the information. Previous studies show that the type of information can affect stated preferences and the validity of value estimates, but how respondents process this information has been less researched. Our study is the first to exogenously manipulate factors of engagement in a stated preference survey. Drawing on stated preference guidance and psychological concepts, we estimate effects of (1) quiz questions and (2) self-reference questions on (1) engagement, (2) information recall, and (3) stated preferences in a choice experiment survey valuing the ecosystem services of urban gardens in the German cities of Berlin and Stuttgart. Our results indicate that respondents spend more time on the information page when confronted with quiz rather than self-reference questions. For both question types, we do not find effects on recall or stated preferences. Our results are consistent with the previously found relation between respondent engagement and stated preferences being endogenous to respondent characteristics. Furthermore, the results suggest that questions increasing engagement with survey information may affect information processing less than previously assumed, highlighting the importance of testing the understanding and recall of important information in pretests