[Statistical concepts related to negative clinical results]

Zhong Xi Yi Jie He Xue Bao. 2011 May;9(5):487-90.
[Article in Chinese]

Abstract

It is a trend that clinical researchers overemphasize on hypothesis testing and the P value. However, P value has some limitations in the presentation of study results. Power is a tool for study design and results explanation, especially for negative results. Confidence interval contains more information for clinical results than P value. P value, power, and 95% confidence interval should be provided in the clinical reports to facilitate the interpretation of the results.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Confidence Intervals
  • Data Interpretation, Statistical*
  • Outcome Assessment, Health Care*
  • Research Design*