IC Value
Search
About Us
Authors
Reviewers
Editorial Board
Contact Us

Determining the significance of the research results beyond p value

Authors:Dinesh Kumar Bagga
Int J Biol Med Res. 2018; 9(2): 6278-6281  |  PDF File

Abstract

The p value or statistical significance is of limited value from a clinical perspective because, a research finding with statistical significance may have little clinical significance whereas, a research finding with no statistical significance may be of clinical significance. Therefore, the statistical significance testing alone limits understanding and applicability of research findings in clinical decision‐making due to its inability to provide all information about the clinical significance. While most researchers focus on statistical significance, clinicians and clinical researchers need to focus on clinical significance beyond statistical significance. Evidence‐based practice is supposed to impact clinical decision‐making in clinical practice based on proper interpretation of clinical research and implementation of clinically significant research outcome rather than statistically significant research outcome. A published journal article with a statistically significant result does not necessarily make it important as the statistically significant result does not mean that the difference between groups is big enough to be clinically significant or meaningful. Rather it simply means whether the difference is real (not explainable on sampling variability) or attributable to chance. Therefore, it is rather more important to understand what the p value is not than what it is. By understanding the limitations in clinical research posed by the heterogeneity of patient samples, small sample sizes, and limitations of hypothesis testing itself, clinicians must include other clinically‐relevant measures such as effect size, clinically meaningful differences, confidence intervals, and magnitude‐based inferences. Clinical researchers need to present research results with clinical significance beyond statistical significance and clinicians need to interpret and implement those clinically significant outcomes in their evidence based approach to clinical decision-making.