Thu 4 April 2019

How to read an ANOVA summary in R

Written by Hongjinn Park in Articles

Let's say you want to test three different gasolines to see which yields the best MPG. You get five identical cars and have them drive the same course until they run out of gas. You do this for each gasoline. Probably not good for the environment. Assuming the randomness is Normally distributed, an ANOVA model can be used to assess which gasoline is the best.


After running ANOVA, R provides a lot of useful output. I made my own "legend" on what the output means. Check it out here.

The $p$-value is used to assess the hypothesis that the means are all the same. This is done by calculating variance in two different manners and if there is little difference, then it's more likely that the means are all the same.



Articles

Personal notes I've written over the years.