For more information.
For questions, please email Shayne Trotman (shayne.trotman@nyu.edu)
Abstract:
We shed new light on replications of laboratory experiments in economics and recall largely ignored methodological problems in the current praxis of analyzing experimental data. The basis of our investigation is a recent study by Camerer et al. (2016). We point out that most of the papers therein refer to a treatment effect that cannot be identified with the used test. We reinvestigate the original data sets and the replicated ones using tests that uncover treatment effects and thereby present a novel methodology for designing replications and establishing replicability.