Are P values given more importance than they deserve?

Abstract

In many of the scientific reports, the conclusions are based on the P values. The results are dichotomized based on the cutoff of P values. Often, the clinical significance is ignored over statistical significance. The sole dependence on P values can lead to erroneous conclusions. The confidence intervals in turn can provide more useful information. They provide applicability of the data in whole populations. This article provides insight into P values, the null hypothesis, and confidence intervals.

Keywords

• confidence interval
• p value
• statistical significance