I have written a new paper about the p-value controversy. Here is the
abstract:
_________________________________
The editorial in the March 2019 special issue of THE AMERICAN STATISTICIAN
(TAS) promises several useful outcomes if we abandon using the concept of
"statistical significance" in scientific research. Two of the promises are
that (a) abandoning statistical significance will lead to fewer
false-positive errors in scientific research, and (b) abandoning statistical
significance will make it easier to replicate scientific research results.
The present paper discusses the role of statistical significance as a
gateway to publication in scientific journals. The paper then shows how
abandoning statistical significance as a gateway to publication will lead to
MORE false-positive errors in published research and will make published
research results HARDER to replicate.
Keywords: Hypothesis test; p-value; False-negative error; Scientific
publishing
_________________________________
The paper is at https://matstat.com/macnaughton2019a.pdf
Donald Macnaughton
[log in to unmask]
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|