Suchergebnisse

ABSTRACTThis article considers how a key legislative vote—that is, the August 2011 vote to raise the federal debt ceiling—influenced the 2012 elections for the US House of Representatives. Two outcomes are analyzed: (1) the incumbents' ability to retain their seats through the 2012 general election, and (2) their share of the two-party vote for members who faced a general-election competitor. In developing this study, the research design was registered and released publicly before the votes were counted in 2012. Therefore, this article also illustrates how study preregistration can work in practice for political science. The findings show that seat retention did not vary with the treatment; however, incumbents who voted against raising the debt ceiling earned an additional 2.4 percentage points of the two-party vote.

ABSTRACTThis article describes the current debate on the practice of preregistration in political science—that is, publicly releasing a research design before observing outcome data. The case in favor of preregistration maintains that it can restrain four potential causes of publication bias, clearly distinguish deductive and inductive studies, add transparency regarding a researcher's motivation, and liberate researchers who may be pressured to find specific results. Concerns about preregistration maintain that it is less suitable for the study of historical data, could reduce data exploration, may not allow for contextual problems that emerge in field research, and may increase the difficulty of finding true positive results. This article makes the case that these concerns can be addressed in preregistered studies, and it offers advice to those who would like to pursue study registration in their own work.