Chapter 10 resources - Goodness of fit and analysis of variance

Hypothesis testing glossary

The University of Lancaster in England has developed an online glossary, for easy lookup of many of the terms used in this class. It is a good reference or check to make sure whether you know a definition.

Another dictionary of statistical terminology. This one, from Montana State University is down home and funky, but complete.

Goodness of fit and categorical independence
The Chi-squared distribution can be used to see whether the disribution of some random variable fits a theoretical probability distribution (e.g. normal distribution). Here is a brief introduction to this test.

The engineering statistics manual has a pretty clear explanation of the use of the Chi-square distribution for goodness of fit. It also gives several examples and shows how the Chi-square compares to other, nonparametric (no assumption of a probability distribution) methods.

David Lane gives some good links to instructional demos on the use of Chi-square for goodness of fit and categorical independence in a 2x2 contingency table. There is also a tutorial on Chi square in hyperstat which is on this link.

Analysis of Variance

There are many tutorials on ANOVA. This by David Humphreys is pretty straightforward.
ANOVA was developed by Ronald Fisher. He also codified many procedures in statistics, making a science out of an art.
This analysis of variance applet allows you to put in your data and does ANOVA tests on them. Also, there are examples of data sets for one-way and two-way ANOVA which you can do the tests on.
Here's a multiple regression tutorial for the TI-83.
The history of statistics is outlined in this web page from UCLA. This includes Francis Galton's (1886) first use of regression.
A more detailed description of how regression was developed by Galton and Pearson. It shows how much we take for granted now knowing the relationship of the normal and t distributions.

Web-based calculators

A t-distribution calculator I like gives and shows you the area under the t between any two points. It has slider bars for changing the boundaries and the degrees of freedom. It is pretty easy to use.

A chisquare-distribution calculator I like gives and shows you the area under the chisquare between any two points. It has slider bars for changing the boundaries and the degrees of freedom. It is pretty easy to use.

There is an F-distribution calculator. You give it degrees of freedom of numerator and denominator and p value and it gives the f value such that p fraction of the area is to the left of that f value. Or you can give it F and it gives you the fraction of the area to the left of it..