>Episode 2: Hardy and Weinberg to the rescue
Picking a good statistical test
The next morning, the ME called the Assistant into his office. He had another thick official looking report in his hand and the Assistant knew this couldn't be good.
“Some very expensive genetic tests on 200 people have shown that the proportions of unaffected, carrier, blue phenotypes are 62%, 36%, and 2%, whereas you just predicted 61%, 34%, and 5%. Care to explain yourself?”
What kind of test could the Assistant use to show that the study's distribution of genotypes (62:36:2) does, in fact, fit his model (61:34:5), and that the difference is just due to random 'slop'?
|analysis of variance|
Copyright University of Maryland, 2007
You may link to this site for educational purposes.
Please do not copy without permission
requests/questions/feedback email: firstname.lastname@example.org