Logo Logo
Help
Contact
Switch Language to German
Imai, Taisuke; Zemlianova, Klavdia; Kotecha, Nikhil; Camerer, Colin F. (18. September 2017): How Common are False Positives in Laboratory Economics Experiments? Evidence from the P-Curve Method.
Full text not available from 'Open Access LMU'.

Abstract

Scientific conclusions depend on how data are collected and analyzed, and whether norms and editorial practices suppress false positives. Poor replicability of results in some areas of medicine and psychology have raised concerns about how widespread such bad results might be in other areas of science. We analyzed all laboratory experiments in economics published in seven leading journals between 2009 and 2016 using a p-curve, which is the frequencies of reported p-values in equal-sized bins, spanning relatively strong (p<.01) and marginal (.04<p<.05) results. The observed p-curve is strongly right-skewed. P-hacking does not appear to be common in laboratory experimental economics.