Most Neuroscience Data Is Not Normally Distributed: Analyzing Your Data in a Non-normal World

While the most common statistical tests assume that the error of the dependent variable follows a normal distribution, dependent variables in translational neuroscience studies often fail to meet this assumption. Common statistical tests like the t test and ANOVA are based on the normality assumption, but quite often these tests are used without checking whether the dependent variable meets the normality assumption which can lead to erroneous interpretations and conclusions about observed associations. There is a significant need for the neuroscience community to utilize nonparametric statistics, particularly for regression analyses. Neuroscientists can greatly enhance the rigor of their analyses by understanding and utilizing nonparametric regression techniques that provide robust estimates of associations when data are skewed. This commentary will discuss and demonstrate analytic techniques that can be used when data do not meet the assumption of normality.