When Studies Are in Error: Basic Statistical Vocabulary Needed to Understand Clinical Studies
Martin A . Weinstock
Background: Accurate
understanding of certain basic statistical terms and principles
is key to critical appraisal of published literature.
Objective: This review describes type I
error, type II error, null hypothesis, p value, statistical
significance, alpha, two-tailed and one-tailed test, effect size,
alternate hypothesis, statistical power, beta, publication bias,
confidence interval, standard error, and standard deviation,
while including examples from reports of dermatologic studies.
Conclusion: The application of the
results of published studies to individual patients should be
informed by an understanding of certain basic statistical
concepts.
Full text available in the print edition / Pour le texte intégral veuillez consulter la version imprimée.
