Reporting statistics

Deze pagina lezen in het Nederlands.

Author: dr. ir. N. van Geloven
Authorship

This page was translated from the Dutch-language original using AI and has not been fully reviewed. Where the two differ, the Dutch version takes precedence.

Contents

• How do I report the results of non-significant findings?
• How do I report very small p-values?
• Guidelines
• Additional resources

How do I report the results of non-significant findings?

We are currently writing up a study in which we used logistic regression. There is some uncertainty about how best to record these outcomes in the final article. It concerns the results with a non-significant outcome, for which we have currently recorded the p-value along with the B value (when significant, we record the OR and confidence interval).

There are still many differences in how the results of statistical analyses are presented. You could take the following approach: if you make a table with the results of your analyses in it, list all tested variables — including those that were not significant. This clearly shows what was in your model, and anyone who later wants to investigate those non-significant factors can see exactly what came out of your analysis. For a logistic regression, stating the combination of OR - confidence interval - p-value is often sufficient. The B is directly related to the OR (OR=e^B, the caret means "to the power of"), so from one you can always calculate the other yourself. If you also discuss the results from the table in the text, you can cite the significant ones along with the number and the non-significant ones without numbers (these can be looked up in the table).

Do note: some journals are very strict about overlapping information between the table and the text. In that case, you choose one of the two approaches. In any case, I would make sure that the OR/conf int/p-value of non-significant results are mentioned somewhere as well.

How do I report very small p-values?

I received the following question from a reviewer:

Statistics are reported to be p=0.00 in several instances. Usually one chooses a priori the risk of a type I error by deciding that p < 0.05 will be considered statistically significant. Some authors report the actual p-value if less than 0.05 to signify a strong probability that the null hypothesis is wrong. If p-values are to be reported, the necessary number of digits must be reported, e.g. p<0.001 or similar, but the p-value should not be abbreviated to 0.00 unless the test actually returns this value, highly unlikely.

What is the best thing for me to report?

In general, rounding is fine, but if the result then comes out as 0.00, it is unclear whether this is, for example, 0.004 or 0.0000000000000000000000000000001. The best approach is to report such a number with more digits after the decimal point (0.004), or, for example, with <0.001 when there are still zeros after 3 decimal places.

Guidelines

• The SAMPL Guidelines. Basic Statistical Reporting for Articles Published in Biomedical Journals: The “Statistical Analyses and Methods in the Published Literature” or The SAMPL Guidelines
• CONSORT statement The CONsolidated Standards of Reporting Trials Statement is intended to improve the reporting of a randomized controlled trial (RCT), enabling readers to understand a trial's design, conduct, analysis and interpretation, and to assess the validity of its results
• EMA Scientific guideline ICH Topic E9 Statistical Principles for Clinical Trials
• Statisticians in the Pharmaceutical Industry: Guidelines For Standard Operating Procedures for Good Statistical Practice in Clinical Research - Guidelines for conducting and reporting statistical analyses according to GCP standards.

Additional resources

• The Equator Network offers all the latest versions of guidelines associated with specific designs (meta-analysis, diagnostic study, observational study, etc.)
• Writing the Empirical Journal Article, Daryl J. Bem - Many useful pieces of advice about what should be included in an article. Also advice on language/form and on responding to reviewer comments.
• Gardner MJ, Machin D, Campbell MJ, “Use of Check Lists in Assessing the Statistical Content of Medical Studies” BMJ. 1986;292:810-2
• Bailar JC, Mosteller F, “Guidelines for Statistical Reporting in Articles for Medical Journals” Ann Intern Med. 1988;108:226-73

---

Finished reading? You can go to the Overview of all topics, or to Choosing a test for help picking a suitable test or analysis. Want to learn more about biostatistics? Then follow the AMC e-learning Practical Biostatistics. Did you find something on these pages that isn't correct? Does a link not work? Or would you like to contribute to the wiki? Then get in touch with us.

This biostatistics wiki is an initiative of the former statistics helpdesk of Amsterdam UMC, location AMC. Amsterdam UMC staff can request statistical support via intranet.