Never P alone: The value of estimates and confidence intervals

Tom Lang

Volume 25, Issue 3 - Statistics

Never P alone: The value of estimates and confidence intervals

Author: Tom Lang

Abstract

How results are reported influences how they are interpreted. Although P values have been granted great importance, they have no clinical interpretation. Rather, they are a measure of chance as an explanation for the results. Their either or interpretation takes attention away from the results themselves – the difference between groups or the effect size–which are more important. Effect sizes are also estimates. Estimates are only useful if they are accompanied by a measure of precision. In medicine, this measure is usually the 95% confidence interval (CI). This article explains the concepts underlying CIs and illustrates how they are more useful than P values in reporting research. As such, journals are increasingly asking for CIs, instead of, or at least in addition to, P values.

Download the full article

References

Rowntree D. Statistics without tears: an introduction for nonmathematicians. London: Penguin Books; 2000.
Lang TA, Secic M. How to report statistics in medicine: annotated guidelines for authors, editors, and reviewers. 2nd ed. Philadelphia: American College of Physicians; 2006.

Volume 25, Issue 3 - Statistics

Never P alone: The value of estimates and confidence intervals

Author: Tom Lang

Abstract

References

Search

Articles

Member Login

Links

Editoral Board