Statistical analysis comparing two groups is most frequently reported in the literature; an unpaired t-test is popular to inform results from two independent groups. The t-test is parametric and makes certain assumptions about the data. Literature, however, has shown that the t-test is robust to statistical assumptions violation. Many researchers, therefore, do not test assumptions to apply t-test. The robustness property of the t-test has certain limitations. A researcher interested in reporting results using a t-test must know the following:
• What are the assumptions for an unpaired t-test?
• Are assumptions essential and need to be tested?
• What is robustness in the context of statistical tests?
• Are there any reporting standards to report t-test results?
The t tests | The BMJ [Internet]. The BMJ | The BMJ: leading general medical journal. Research. Education. Comment. 2020 [cited 2022 Jun 9]. Available from: https://www.bmj.com/about-bmj/resources-readers/publications/statistics-square-one/7-t-tests
Zimmerman DW. Comparative power of student t test and Mann-Whitney U test for unequal sample sizes and variances. J Exp Educ 1987;55(3):171–174. DOI: 10.1080/00220973.1987.10806451
Kishore K, Jaswal V. Statistics corner: data management in Rcmdr. J Postgrad Med Educ Res 2022;56(2):102–105. DOI: 10.5005/jp-journals-10028-1571
Altman DG, Bland JM. Absence of evidence is not evidence of absence. BMJ 1995;311(7003):485. DOI: 10.1136/bmj.311.7003.485