Data science in medical research

Research output: Contribution to journalArticlepeer-review


Background and Objectives: A common technique employed in medical research is to answer questions using statistical infer-ence and to calculate an appropriate test statistic and confidence interval that will determine whether some hypothesis should be rejected or otherwise. There are many issues to consider, including the controversial matter of whether a one-or two-sided alternative is appropriate. The objective of this paper is to provide a precise interpretation of a confidence interval as well examining how conditional probability should be interpreted.

Methods and Materials: A novel approach is undertaken for explaining the p-value in relation to a confidence interval that has been used to test a medical hypothesis. It is also apparent that even skilled medical researchers have not been sufficiently trained in statistical analysis do not have the ability to interpret statistical results. Examples to illustrate the techniques are pro-vided.

Results: The results in this paper will enable the medial researcher to better understand the nature of confidence intervals and to use them more effectively in reaching conclusions.

Conclusion: It is known that there are a number of medical research articles appearing in refereed journals that contain erroneous conclusions through the misinterpretation of data analysis. This paper uses a case study to illustrate that their findings can often be misleading or just plain wrong.

Original languageEnglish
Pages (from-to)650-653
Number of pages4
JournalInternational Medical Journal
Issue number6
Publication statusPublished - Dec 2021


  • Bayes Theorem
  • Clinical trials
  • Confidence interval
  • Hypothesis testing


Dive into the research topics of 'Data science in medical research'. Together they form a unique fingerprint.

Cite this