ABSTRACT
Surgery, like any other scientific discipline, requires a systematic analysis of all its different variables in order to prove the real significance of research findings. Statistics, the science of numerical evaluation, can thoroughly help to determine the real value of surgical treatment. In this work, we study the statistical tests and principles needed to demonstrate their role in surgical research.
Without a strong statistical background, a researcher may feel overwhelmed when deciding what statistical methods to utilize in research. Determining what type of data to collect and what hypothesis test to run can alter the entire way a surgical study is conducted.
The relationship between power, sample size and effect size is discussed as well as the components necessary for a power analysis. Selecting an appropriate sample size is of utmost importance in any type of research since an undersized sample can invalidate an entire study.
Categorical surgical data, numerical data, and the appropriate statistical procedures needed for analysis are reviewed. Methods discussed include the 2-Sample t-test, Mann Whitney U test, Kruskal-Wallis, ANOVA, Chi-Square test and Fisher's exact test.
Notes
1 All statistical tests run on Minitab 16 software.