Hello, I have a question on the t-test formula for more than 2 independent samples, variances unknown, but assumed to be equal.
It is stated that the degrees of freedom is equal to the total sample size minus the number of variables.
So, if I have 4 samples instead of 2, would the formula be “nw + nx +ny + nz – 4”?
Is there a maximum limit of sample size for the t-test in this case? I’m aware that the sample should be below 30 (or 40?) otherwise we should use z. So for example, if I want to compare apple prices in NY, MI, LA, TX, WY, and there are 6 price data for each state, which adds up to 30. Is this the same as having 10 apples for NY, LA, TX (3 states, which adds up to 30 as well)?
When you work with more than two samples, you don’t apply the normal hypothesis testing as illustrated in the course. You should apply another technique called ANOVA (look for it on the net), and it’s outside of scope of the course,
I hope that I was clear 😀
Hi Jaber, ah I see! OK, I will look for ANOVA. Thanks 😀