Rumors of the 'Nonreplication' of the 'Motivated Numeracy Effect' Are Greatly Exaggerated
15 Pages Posted: 31 Aug 2017 Last revised: 30 Oct 2017
Date Written: August 26, 2017
Abstract
This paper does three things. First, it describes the design defects (principally, the lack of statistical power) that make it misleading for Ballarini & Sloman (2017) to claim that they “failed to replicate” the results of Kahan, Peters et al. (2017). Second, it presents the positive results of our own replication study. Finally, we conclude with a brief discussion of why confining assertions of non-replication to studies that satisfy emerging replication protocols—in particular the imperative of “faithful recreation of a study with high statistical power” (Brandt, Ijzerman et al 2014, p. 217)—is essential to the contribution such studies can make as building blocks of a cumulative science.
Keywords: replication, gun control, risk perception, science communication
Suggested Citation: Suggested Citation