NOBEL PRIZE WINNER IN ECONOMICS RETRACTS HIS ENDORSEMENT OF SOME STATISTICAL STUDIES. One of the changes that has occurred in the last ten years in the use of statistics is the increased emphasis on whether a study can be replicated. There are a number of journals and websites now devoted to attempts at replication. One of these websites is Retraction Watch, which can be found at retractionwatch.com. This post on the Retraction Watch blog describes how Daniel Kahneman, a Nobel Prize winner in Economics, acknowledged that he placed too much faith in weak studies in one chapter of his book THINKING FAST AND SLOW.
The studies are said to be “underpowered”—- a short translation of “underpowered” is that the studies are based on too small a sample to support the conclusions they reach.
Kahneman points out the irony that the first paper he published, back in 1971, was about what he and his co-author Amos Tversky, called the “’law of small numbers’, which in Kahneman’s words,”allows researchers to trust the results of underpowered studies with unreasonably small samples”.
The irony shows how omnipresent has been the use of samples that are too small, so that even a major scientist who has previously published on the problem of small sample sizes could forget the issue.