The Ultimate Irrelevance Proposition in Finance?
George Andrew Karolyi
Cornell University - Johnson Graduate School of Management
April 13, 2011
I survey 457 published papers in top Finance journals across two decades to assess whether they misuse tests of significance. Over 80% of published studies are diligent about distinguishing between statistical and economic significance and about quantifying and interpreting the economic magnitudes of the statistical relationships they measure. Yet, only 10% of them acknowledge limits to the power of their tests and fewer still do anything about them. Recent demographic trends in publishing such as larger co-author teams, the increased participation by non-North American scholars, by women and by those outside the top Finance departments are not associated with these outcomes.
Number of Pages in PDF File: 40
Keywords: Empirical methods in Finance, significance tests, power of the test
JEL Classification: G1, C1, C58
Date posted: April 16, 2011