Advertisement

Scientometrics

, Volume 116, Issue 3, pp 2193–2194 | Cite as

Response to commentary on “Is NHST logically flawed”

  • Jesper W. SchneiderEmail author
Article

I thank the Editor for giving me the opportunity to respond to Dr. Patriota’s (2018) brief comment in relation to two of my pervious publications on null hypothesis significance testing (NHST) in this journal (Schneider 2015, 2018). I shall be brief myself in my response.

If I understand Patriota correctly, he insists that I am mistaken when I describe p values as conditional probabilities. According to Patriota, from a frequentist point of view this is incorrect because p values supposedly do not “respect the basic properties of conditional probabilities” (Patriota 2018, p. 2).

I am familiar with Patriota’s claim, one he shares with other frequentists, see for example Larry Wasserman’s blog entry but also the critical comments accompanying it,1 as well as the parallel discussion on Andrew Gelman’s blog;2 Patriota was an avid debater in both discussions. But like Gelman and several other commenters in the debate, I do not think I have misunderstood anything. Whether the pvalue is...

References

  1. Briggs, W. M. (2017). The substitute for p-values. Journal of the American Statistical Association, 112(519), 897–898.MathSciNetCrossRefGoogle Scholar
  2. Fisher, R. A. (1956). Statistical methods and statistical inference. Edinburgh: Oliver and Boyed.zbMATHGoogle Scholar
  3. Hacking, I. (1965). The logic of statistical inference. Cambridge: Cambridge University Press.CrossRefzbMATHGoogle Scholar
  4. Howson, C., & Urbach, P. (2006). Scientific reasoning. The Bayesian approach (3rd ed.). Peru, Ill: Open Court.Google Scholar
  5. Neyman, J. (1937). Outline of a theory of statistical estimation based on the classical theory of probability. Philosophical Transactions of the Royal Society London, Series A, 236, 333–380.CrossRefzbMATHGoogle Scholar
  6. Neyman, J., & Pearson, E. S. (1933). On the problem of the most efficient tests of statistical hypotheses. Philosophical Transactions of the Royal Society London, Series A, 231, 289–333.CrossRefzbMATHGoogle Scholar
  7. Patriota, A. G. (2018). Is NHST logically flawed? Commentary on: “NHST is still logically flawed”. Scientometrics.  https://doi.org/10.1007/s11192-018-2817-4.Google Scholar
  8. Pollard, P., & Richardson, J. T. (1987). On the probability of making type I errors. Psychological Bulletin, 102(1), 159–163.CrossRefGoogle Scholar
  9. Schneider, J. W. (2015). Null hypothesis significance tests. A mix-up of two different theories: The basis for widespread confusion and numerous misinterpretations. Scientometrics, 102(1), 411–432.CrossRefGoogle Scholar
  10. Schneider, J. W. (2018). NHST is still logically flawed. Scientometrics, 115(1), 627–635.CrossRefGoogle Scholar
  11. Sober, E. (2008). Evidence and evoluation. The logic behind science. Cambridge: Cambridge University Press.CrossRefGoogle Scholar

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2018

Authors and Affiliations

  1. 1.Danish Centre for Studies in Research and Research Policy, Department of Political ScienceAarhus UniversityAarhusDenmark

Personalised recommendations