Bimodal Distributions

A statistical distribution that is a non-normal distribution having more than one hump or bell curve, as opposed to just one representing the center where most of the data points should be plotted. In these distributions, the data isn't just biased or skewed, but it appears to have more than one mode. This may occur when the data being measured is actually coming from two separate populations. In these cases, the data should be split in order to reflect their respective population. Although the non-normality of these distributions may appear visually obvious, they should be statistically validated such as in using an Anderson-Darling (AD) test of a Normality Test or Probability Plot.
Related StatStuff Videos: 
   -Section 5-Measure Phase, Lesson #8 - Distributions: Overview
   -Section 5-Measure Phase, Lesson #10 - Distributions: Non-Normal