- Quiz Thursday.
- We're going to polish off Packet 11 today: Two Sample Inferences.
- Then we move on to our last packet, Packet 12: ANOVA (we could say "multi-sample" inference (instead of just "two").
"ANOVA" stands for "Analysis of Variance". Variance is another measure of spread or "variability" (it's just the square of the standard deviation).
We will tend to use the ranges in the graphics to represent spread, and to give us some idea of the "F statistics" calculated.
The F-distribution looks a lot like a $\chi^2$, and also possesses degrees of freedom (but it's even a little more complicated, because there's one for the numerator and one for the denominator of a quotient!).
In this course we consider it more important to gain an intuitive idea behind the F statistic, rather than a computational one.
By the way, the two sample unpaired design we just studied in Packet 11 is just a special case of ANOVA with just two samples. An alternative way of saying that is that ANOVA is a generalization of the unpaired sample inferences for the mean.
- The Standard Normal Table of probabilities.
- Statcrunch