Representative, cross-national values surveys like the World Values Survey and Arab Barometer provide social scientists with the ability to test hypotheses related to values and opinions on a global scale. Unfortunately, while survey methodology has expanded the scope of scientific understanding of these phenomena, this has come at the expense of depth. In the context of the Arab world, for example, survey research shows that most Arabs express political orientations drawing on both democratic and Islamist ideologies, but very few studies have investigated the meanings of the survey items – including concepts like shari’a and democracy – to individual respondents. This project proposes to overcome this problem using cognitive interviewing, which involves asking respondents to answer a set of survey items along with open-ended follow-up questions about their answers. I use this technique to investigate the relationship between support for shari’a and support for democracy based on one-on-one interviews with fifty Tunisians performed between August 2013 and March 2014. Results indicate that relying on responses to survey questions as straightforward indicators of support for either shari’a or democracy is inappropriate. Responses to questions about implementing shari’a may have more to do with attitudes about literalist religious movements than attitudes about secular or religious governance, and responses to questions about democracy may have more to do with attitudes about the west than attitudes about democratic practices as such.