On not testing the foreign-language effect: A comment on Costa, Foucart, Arnon, Aparici, and Apesteguia (2014)

Florian Wickelmaier
DOI: https://doi.org/10.48550/arXiv.1506.07727
2015-06-25
Applications
Abstract:In their first five studies, Costa, Foucart, Arnon, Aparici, and Apesteguia (2014) fail to provide a statistical test of the foreign-language effect. Instead, the authors employ a procedure in which they test the framing effects separately for the native and the foreign language conditions. Such a procedure, however, is inappropriate when comparing two effects; rather, a test of their difference is required. Using the original data, it is shown that in four out of the five studies the authors' conclusions about the existence of a foreign-language effect are invalid.
What problem does this paper attempt to address?