Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
January 16, 2018Anesthesia & AnalgesiaOpen Access

Statistical Significance Versus Clinical Importance of Observed Effect Sizes: What Do P Values and Confidence Intervals Really Represent?

View Full Paper
Ask AI
Bookmark
Share

Authors

PSPatrick SchoberGeneral CardiologySBSebastiaan M. BossersVrije Universiteit AmsterdamLSLothar A. SchwarteDutch Cancer Society

Discussion

Loading...

Member takes

Implication

Tutorial review demonstrates the clinical interpretation of effect sizes and confidence intervals beyond p values, highlighting the need to evaluate practical importance alongside significance.

Key Points

  • Explain the essential distinction between statistical significance and clinical importance while providing intuitive guidance on interpreting effect sizes, confidence intervals, and p-values.
  • Tutorial review evaluating more than 70 effect size measures, including standardized mean differences, risk ratios, odds ratios, and correlations.
  • Conceptual analysis illustrating how confidence intervals quantify uncertainty and how p-values function beyond binary significance thresholds.
  • Null hypothesis significance testing and isolated p-values fail to convey the magnitude, direction, or practical clinical relevance of observed treatment effects.
  • Pairing point estimates with confidence intervals reveals the plausible range of effect sizes in the target population, enabling direct assessment of clinical importance.

Cite This Study

Schober et al. (2018) studied this question.

synapsesocial.com/papers/6a0cac4ac63d985fc0886091https://doi.org/10.1213/ane.0000000000002798
View Full Paper
Ask AI
Bookmark
Share