Procedural transparency and the credibility of election surveys

The success and influence of survey-based electoral research is fueling the ambitions of survey analysts and producers. As a result, many new forms and uses of survey data are emerging. These new activities bring with them important questions about credibility. I address several of these questions by discussing common practices in the production and analysis of election surveys. I contend that the continuation of some of these practices threatens the credibility of individual studies and, in some cases, the election-oriented survey enterprise as a whole. In all of these cases, however, I argue that a commitment to increased transparency about analytic and/or production decisions can enhance credibility. In the process, I provide suggestions and examples of how transparency can be increased.