VOLUME 57 , ISSUE 2 ( April-June, 2023 ) > List of Articles
Kamal Kishore, Vidushi Jaswal
Keywords : Chi-squared test, Statistical analysis, Wilcoxon–Mann–Whitney test
Citation Information : Kishore K, Jaswal V. Statistics Corner: Paired Groups. J Postgrad Med Edu Res 2023; 57 (2):90-93.
DOI: 10.5005/jp-journals-10028-1626
License: CC BY-NC 4.0
Published Online: 31-05-2023
Copyright Statement: Copyright © 2023; The Author(s).
Health researchers frequently assess changes in patients’ disease status after giving interventions. Routine independent tests, such as t-tests (continuous outcome), Wilcoxon–Mann–Whitney (nonnormal continuous or ordinal outcome), or Chi-square (nominal outcome), will not be applied here. There are different statistical techniques to analyze data from paired observations. A researcher designed a pre–post study and collected data consisting of continuous and categorical variables. The literature search recommends applying paired t-tests for continuous data. However, the researcher is confused as data is not normally distributed when the normality assumption is assessed for pre and postgroup—the difference (post–pre) is normally distributed. Researchers, therefore, have a few vital questions related to the testing of paired data. What are the assumptions of paired t-tests? What test to apply for ordinal paired data? What test to apply for nominal paired data? What are the null and alternative hypotheses (H1) for paired data?
© Jaypee Brothers Medical Publishers (P) LTD.
