**Principal investigators:** **Adam Levine** Cornell University Email: ASL22@cornell.edu Homepage: http://falcon.arts.cornell.edu/asl22/Home.html **Yanna Krupnikov** Northwestern University Email: yanna.krupnikov@northwestern.edu Homepage: http://faculty.wcas.northwestern.edu/~ykd778/Home.html **Sample size:** 1029 **Field period:** 6/17/2009-10/26/2009 **Abstract** Numerous researchers are interested in the conditions under which people participate in politics. Focal questions include the effect of stimuli such as ads, direct mail, and other information on behaviors such as voter turnout, volunteerism, information acquisition, and donation behavior. While some studies analyze previous behavior, many experiments and surveys focus on people’s willingness to act in the future. In these cases, designing questionnaires that produce valid measures of people's political behavior raises a unique challenge. Should scholars ask people to express their preferences for action or should they ask people to reveal them? While expressed preference measures are often easier to implement, they may result in severe overreporting. Revealed preference measures may offer a more valid measure of people's willingness to act, but can be difficult, if not impossible, to implement in a survey setting. Given these difficulties, in this project we propose an alternative measurement approach based on the psychology of scripts. Our approach leverages the benefits of both expressed and revealed preferences, and in doing so yields an improved measure of people’s willingness to act. **Hypotheses** People will express a lower willingness to act after they consider an imagination script than when they have not considered one. The preferences that people reveal after considering an imagination script will match the preferences that they reveal without the imagination script. The preferences that people express after considering an imagination script will match the preferences that they reveal without the imagination script. **Experimental Manipulations** The experiment involved two factors: the information people received about a real organization, and whether they were asked to express or reveal their preference for volunteering time with that organization. The information they received was either presented as basic information that identified the goals of the organization and its volunteer listserv, or as an imagination script that also talked about the goals and its volunteer listserv, but was framed in such a way that people were asked to imagine being on it. The information was balanced in terms of the language used and the number of words. **Outcomes** Half of subjects were asked to express their preferences by indicating how likely they would be, if asked, to sign up for a listserv to hear about volunteer opportunities with the organization. The other half of subjects were asked to reveal their preferences by actually deciding whether or not to sign up for the listserv. **Summary of Results** Our findings are consistent with our hypotheses. People who received the script expressed a lower willingness to sign up than those who did not, suggesting that the script was bringing to mind considerations that were not present with a standard expressed preference measure. The same difference was not observed among those who were asked to reveal their preferences. In this case, people’s likelihood of actually signing up (i.e. revealing a willingness to sign up) was the same regardless of whether they received the script or not. Finally, the preferences that people expressed after receiving an imagination script matched those of people who were asked to reveal their preferences without the script, suggesting that the imagination script was calling to mind the same considerations that were present when people revealed preferences. **References** Levine, Adam and Yanna Krupnikov. 2010. "Measuring People's Willingness to Engage in Political Action." In *Sourcebook for Political Communication Research: Methods: Measures, and Analytical Techniques*, edited by Erik. P. Bucy and R. Lance Holbert (Eds.). New York: Routledge.