What assumptions does an F-test make?
1 Answer
An F-test assumes that data are normally distributed and that samples are independent from one another.
Explanation:
An F-test assumes that data are normally distributed and that samples are independent from one another.
Data that differs from the normal distribution could be due to a few reasons. The data could be skewed or the sample size could be too small to reach a normal distribution. Regardless the reason, F-tests assume a normal distribution and will result in inaccurate results if the data differs significantly from this distribution.
F-tests also assume that data points are independent from one another. For example, you are studying a population of giraffes and you want to know how body size and sex are related. You find that females are larger than males, but you didn't take into consideration that substantially more of the adults in the population are female than male. Thus, in your dataset, sex is not independent from age.
