This project has moved. For the latest updates, please go here.

Please help me to find the correct classes and methods for this simple goodness of fit tests

Sep 11, 2014 at 7:58 PM
Edited Sep 13, 2014 at 5:11 AM
Warning serious NOOB alert

Say, I have a die that is 19 times as likely to produce a six as a one, because it has been tampered with.
When I throw this die 60 times the expected versus observed frequencies of the six possible outcomes are:

1: 10, 1<br/>
2: 10, 10<br/>
3: 10, 10<br/>
4: 10, 10<br/>
5: 10, 10<br/>
6: 10: 19

I want to feed this expected-observed pairs to an algorithm to determine how likely it is that the die indeed has been tampered with.

When I enter the value pairs on this website it calculates a Chi square value of 16.2 and a P value of 0.00629567, indicating that it is very unlikely that the observed results are in accordance with the expected distrbution of the values one to six.

I want to calculate the P value using numerics, but although I can find a ChiSquared class I cannot find how to feed the expected-observed value pairs to it in order te get the P value.

How can it be done?

Thanks for you patience beforehand :-)