European Journal of Biomedical and Pharmaceutical Sciences

SIGNIFICANCE OF CORRECTION OF P VALUE IN STATISTICAL COMPARISONS AMONG MULTIPLE GROUPS IN BIOLOGICAL RESEARCH

Shashi Sharma, Atul Juneja, Tulsi Adhikari, Apoorva Anand

ABSTRACT

The statistical comparisons between two groups among multiple groups are a common phenomenon in biomedical research including clinical trials, observational studies and more importantly genetic studies. It is important to have an idea about interpreting the p value or type 1 error in accepting or rejecting the null hypothesis when there are only two groups vis a vis when there are more than two groups and two groups of interest are being compared. Situation is most commonly encountered in genetic studies when multiple genes are being studied. Such situation of comparing sub sets from the multiple groups demands correcting the p value which is technically known as Family Wise Error rate. This communication presents the concept of this correction particularly to the medical or biological researchers when post hoc test are being employed. This would help in the better appreciation of the results through p values with due consideration of statistical and biological/medical and its importance when sub set from the multiple groups are being compared.

Keywords: P value, multiple comparisons, Post Hoc tests family wise correction error rate.