P-Value

Definition

The p-value is the probability of observing a test statistic as extreme as, or more extreme than, the one calculated from your data, assuming the null hypothesis is true. A small p-value (typically below 0.05) is taken as evidence against the null hypothesis.

Why It Matters

The p-value is the most widely reported measure of statistical evidence in published research, yet it is among the most misunderstood. It does not tell you the probability that the null hypothesis is true, nor the probability that your result is a fluke. It simply measures how compatible your data are with the null hypothesis. Understanding this distinction is essential for interpreting results correctly and avoiding common misinterpretations that can undermine research credibility.

Example

Suppose you test whether a new teaching method improves exam scores. You obtain a mean difference of 8 points and a p-value of 0.03. This means that if the teaching method had no real effect, there would be only a 3% chance of observing a difference this large or larger purely by random sampling. You would reject the null hypothesis at the 5% significance level, but the p-value does not tell you how large or important the 8-point improvement is.

Related Terms

• Null Hypothesis
• Statistical Significance
• Confidence Interval (provides more information than a p-value alone)
• Effect Size (quantifies practical importance)
• Type I Error

Software Notes

• SPSS: P-values appear automatically in the output tables for t-tests, ANOVA, regression, and chi-square tests under the "Sig." column.
• R: Most test functions return a p-value in the output object (e.g., t.test(), aov(), chisq.test()). Extract with summary(model) or model$p.value.
• Stata: P-values appear in the regression or test output tables under the "P>|t|" or "Prob > F" columns.