. Q 1. I’ve just recently become familiar with the idea of calculating p-values but am not confident about how to express the results of my hypothesis tests in written form within the Discussion or Conclusions section of a report or manuscript.
A. Please refer to the advice provided on slides 16 to 17 of the following presentation:
Principles of statistical inference: hypothesis testing
. Q 2. I would like to gain a better understanding of the notion of statistical effect size and its importance in reporting my statistical findings relative to the p-value. Where can I gain support with the relevant principles and some clinically contextualized examples?
A. Please refer to
Using Effect Size—or Why the P Value Is Not Enough
by Gail M. Sullivan, MD, MPH and Richard Feinn, PhD
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3444174/
. Q 3. Where can I learn more about the place o confidence intervals when reporting statistical findings?
A. You should find the following reference useful:
lConfidence intervals rather than P values: estimation rather than hypothesis testing.
M J Gardner and D G Altman
. Q 4. Where can I obtain structured guidance on how to present the findings and conclusions from statistical hypothesis tests for specific procedures in an appropriate way?
A. You are strongly advised to consult the e-book How to Report Statistics in Medicine.
If you are registered with the University of Edinburgh, you can consult the electronic version of this book via the University’s library discovery system, DiscoverEd. I have ordered and selected this version of the book specifically to support undergraduate medical students in improving their knowledge of what is appropriate to report from a range of statistical output arising from different types of statistical test. I have also highlighted it for many years in the University of Edinburgh Year 4 (now Year 5) Student Selected Component a (SSC4a (now SSC5a)) handbook.
Here are some reference details:
-
- Title:
How to report statistics in medicine annotated guidelines for authors, editors, and reviewers
- Title:
-
- Author: Thomas A. Lang (Thomas Allen)
-
- Michelle Secic 1967-
-
- Publisher: New York : American College of Physicians
-
- Publication Date: 2006], ©2006
-
- Edition: Second edition.
-
- Format: 1 online resource (xxii, 490 pages) : illustrations.
- Language: English.
Also, a key issue of relevance to your query is being much more tentative about the interpretation of findings, that is not over-stating them, particularly where there is a modest sample size. For further learning in this area, you are encouraged to consider the BMJ reference Statistics notes: Absence of evidence is not evidence of absence.
Also, don’t assume that correlation implies causation. The brief article When Correlation is Just for Fun helps to crystalize this point with reference to a concrete example from the New England Journal of Medicine!
Presenting the Findings and Conclusions of Statistical Hypothesis Tests by Margaret MacDougall is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
