Hanif and Ajmal (1) reviewed 80 randomly selected research articles published in indexed local journals of Pakistan. They excluded case series and case reports.
Descriptive statistics was used in 32 (29%) and inferential statistics used in 33 (41%) articles. Wrong statistical analysis was used in 23 (29%) articles. The design of the study was not described in 42 (53%) of the articles.
Almost all, 74 (93%), did not present criteria for sample size estimation or power calculation, and 57 (71%) articles used inferential statistics without mentioning whether the assumptions for the used statistical tests were met or not.
The authors conclude that the standards of medical research need to be upgraded.
- Hanif A, Ajmal T. Statistical Errors in Medical Journals (A Critical Appraisal). Annals of King Edward Medical University, 2011;17:178-182.