Commentary on the use of statistical tests in experimental pharmacological studies

The meaningfulness of the research findings is largely determined by the depth of statistical processing and the quality of their presentation. However, the paucity of high-quality statistical analysis and data presentation poses a significant challenge for maintaining the consistency of published research results. Incorrect processing of primary data can result in distorted conclusions and significantly complicate the generalization of data on a specific research topic. The authors discuss the most frequent and incorrect techniques used in working with initial data and planning an experiment, including issues related to the calculation of sample size, the incorrect use of statistical tests, the identification of numerical statistical parameters and descriptive statistics, and the use of goodness-of-fit and data presentation tests. They also provide recommendations on the appropriateness or inappropriateness of specific methods. Graphic visualization requires that certain principles be followed. For normal distributions, the sample mean is the most practical measure for position. It is a generally accepted practice to show the standard deviation on the graphs as a measure of variability. Confidence intervals are also presented to offer a visual reference. Standard errors may also be represented, however, they are not indicative of data variability, but rather, they are an estimation of the difference between the sample mean and the mean population value. Medians are used to describe non-normal distributions because they are insensitive to outliers. In those specific cases, interquartile ranges (Q1-Q3) can be used as a measure of variability. In the context of large data sets, bar graphs, histograms, box-and-whiskers, and individual values may be considered to enhance the comprehension of visual presentations. The authors strongly advise against the use of tables for displaying large amounts of data, citing the challenges associated with reading such presentations.