but it is not impossible to get away from arbitrary cutoffs...
That 70+ author Benjamin et al. article promoting .005 replace .05 as a
norm has come under extensive criticism for various reasons,
not the least of which is because such a change could worsen publication
bias and P-hacking, and ignores the costs of false negatives, s
o it would effectively distort the literature even worse in the very
fields that have been most damaged by null-hypothesis significance
testing (NHST) - health, medical, psych, and social sciences.
This 80+ author article advocates instead following Neyman's advice to
base alpha on context,
Lakens, D. et al. (88 authors) (2018), Justify Your Alpha: A Response to
Redefine Statistical Significance, Nature Human Behavior, in press,
available at
https://psyarxiv.com/9s3y6Meanwhile others advocate getting rid of alpha levels entirely, and
making sure all studies get into print or at least into online
searchable data bases, e.g.,
Amrhein, V., Korner-Nievergelt, F., and Roth, T. (2017), ???The earth is
flat (p>0.05): significance thresholds and the crisis of unreplicable
research,??? PeerJ 5, e3544.
Amrhein, V., and Greenland, S. (2017), ???Remove, rather than redefine,
statistical significance,??? Nature Human Behavior, 2, 4, doi:
10.1038/s41562-017-0224-0.
see also
Crane, Why Redefining Statistical Significance Will Not Improve
Reproducibility and Could Make the Replication Crisis Worse,
https://psyarxiv.com/bp2z4Greenland, S., Senn, S.J., Rothman, K.J., Carlin, J.C., Poole, C.,
Goodman, S.N. and Altman, D.G. (2016), ???Statistical tests, confidence
intervals, and power: A guide to misinterpretations,??? The American
Statistician, 70, online supplement 1 at
http://amstat.tandfonline.com/doi/suppl/10.1080/00031305.2016.1154108/suppl_file/utas_a_1154108_sm5368.pdf
Original Message------
Check out the response from 60+ statisticians on p-value issue. https://www.nature.com/articles/s41562-017-0189-z This is published in Nature Human Behavior. They recommend using p-value 0.005 instead of 0.05. They make very interesting argument.
Rahul A. Parsa
Iowa State University