What is a Poisson Distribution and what is a Spearman Coefficient?


What is a Poisson Distribution and what is a Spearman Coefficient?

In: 0

A distribution is a fancy way of visualizing the percentage likelihood an event might happen.

There are a number of different types of distributions you might learn about in statistics class.

A Poisson Distribution is a specific type of visualization used to show the percentage likelihood than repeating event is going to repeat within a certain time period. Specifically, you’d use it for events that aren’t interrelated.

For example, if I could visualize the number of cars that will drive by my house in an hour with a Poisson distribution. Each car isn’t directly related to the next, but there is a very likely number, say 20. 2 cars is really unlikely, and 200 is very unlikely. I could say something similar with coin flips that come up heads, or number of defective parts a machine produces etc.

One example of the use of this distribution is it can, for example, prove statistical randomness. If I’m getting a number of defective parts that is wildly at odds with my Poisson Distribution, I can prove something is wrong, for example the events *are related*. Such as the machine is broken thus it’s spitting out tons of defects, or the coin is rigged, or a street closure WAZE is routing local traffic down my street.

The Spearman Coefficient is a value, -1 to 1, that measures how tightly related (or “correlated”) to concepts are. A value of 0 is interpreted as absolutely no relationship, for example the number of legs a duck has and the amount of coffee I drank this morning. 0 to -1 would imply a negative correlation meaning as one goes up, the other down. For example, money in my bank account and amount of groceries I’ve purchased this week. 0 to 1 would be positive correlation or they both move up or down together. Such as hours calories eaten and BMI.

Strength of the correlation would be how close to 0 or -1/1 the coefficient is. Using my examples above calories and BMI is probably only a modest correlation because you have some people who eat a ton of food, but still manage to have low BMIs. The banking example would be a strong negative correlation because every dollar I spend comes out of my bank account.