How should social science researchers deal with data inaccuracies? This article uses Web-based survey data collected from faculty members in three social science disciplines to document variation in views about data editing. Through an analysis of qualitative responses to a hypothetical vignette, the authors demonstrate that a wide range of opinion surrounds the “proper” use of data. Reactions are to some extent contingent on discipline and experience with different types of data and data collection methods. They also depend on characteristics of the data-editing situation—for example, whether the problem is with an independent or dependent variable. Even taking these social and situational factors into account, however, there is still substantial diversity in vignette responses. Normative standards that pervade other aspects of the research process have not yet emerged for data editing.

No abstract