P Value

An often misunderstood term in statistics.

The p-value is a measure of the probability that differences found between groups may be due to random chance fluctuations alone.

A large p-value suggests that while there may be differences between groups, the difference can be attributed to random variations.

A small p-value suggests that the magnitude of the group differences is not likely a result of chance variations alone. The differences cannot be easily explained by randomness. This is taken as evidence in statistical hypothesis testing that the differences suggest the alternative, that the groups have fundamental underlying differences. That is, the differences are statistically significant

Often in clinical studies these differences are related to different treatments. Clinicians when interpreting studies use the p-value to assess statistical significance, and add synthesize this with their own medical training, knowledge, and experience to assess the clinical significance of the study. If deemed appropriate, the clinician may modify his or her clinical practice.