Use and misuse of the p-value

Area of Science:

• Biostatistics
• Clinical Trial Methodology
• Statistical Inference

Background:

• The p-value is a standard metric in clinical research for hypothesis testing.
• Despite its prevalence, common misunderstandings of the p-value persist among researchers.
• Misinterpretation can lead to flawed conclusions in clinical study analysis.

Purpose of the Study:

• To clarify the precise meaning and limitations of the p-value.
• To address prevalent misconceptions associated with p-value interpretation.
• To advocate for the use of confidence intervals over p-values in clinical inference.

Main Methods:

• Review of fundamental statistical inference principles.
• Illustration of p-value concepts and misconceptions using a clinical trial in osteoarthritis.
• Comparative analysis of p-values versus confidence intervals for statistical inference.

Main Results:

• The p-value quantifies evidence against a null hypothesis but does not indicate the probability of the hypothesis being true.
• Common errors include misinterpreting the p-value as the probability of a Type I error or the probability that the null hypothesis is true.
• The clinical trial example highlights how p-values can be misleading without considering effect size and precision.

Conclusions:

• The p-value's interpretation is often oversimplified, leading to significant misconceptions.
• Confidence intervals provide a more comprehensive measure of effect size and precision, enhancing clinical interpretation.
• Confidence intervals are recommended as a superior tool for statistical inference in clinical studies compared to p-values.