There are two peaks in a bimodal distribution. Modes are the peaks of a continuous probability distribution. Above is a graph showing a bimodal distribution. If the peaks are not equal in height, the major mode is at the highest point, while the lower apex represents the minor mode.
Table of Contents
What causes bimodal distributions?
Understanding that your data have a bimodal distribution can help you to better understand your topic. This type distribution often has an explanation. These are few examples.
Combining two processes or populations: Sometimes, bimodal distributions can be created by combining two populations in one dataset. Every underlying conditions have its own mode. Combining them will give you two modes.
Imagine that you are measuring the adult weights of black bears. The graph shows a distribution that has two peaks. The average black bear female weighs 175 pounds, while males average 400. Each average corresponds with a peak in distribution. One peak is for males, the other for females.
Bimodal distributions are often a result of differences between genders. Consider a situation where you are measuring the strength of products on an assembly line. You notice a bimodal distribution. You find out that one shift uses a slightly more complicated procedure, which results in a weaker product. These two processes produce dual peaks.
Natural bimodal distributions
Other cases may show a bimodal distribution of the phenomenon being studied. The size of Weaver Ants and the age at which Hodgkin’s Lymphoma symptoms appear follow a bimodal pattern.
These cases don’t allow you to separate the bimodal distribution from unimodal. Understanding the bimodal nature of your area will allow you to better understand it and identify the most common values near these peaks.
Analyzing Bimodal Distributions
Bimodal distributions are less common, but they’re still important to recognize when they do occur. It is valuable to discover that your data follows a bimodal distribution when you work with mixed populations, conditions or processes. This factor has been identified. It affects the outcome. This variable can be included in future research by you and other scientists.