*** Univariate Chi-squared test (freeware)
*** by Daniel J. Bishop
*** danb2k@hotmail.com

from the Statistics Toolkit, version 3.0
http://www.ticalc.org/archives/files/fileinfo/124/12456.html

The TI-83 has a bivariate chi-squared test built in, but, for some strange reason, there is now univariate chi-squared test.  Unless, of course, you have this program.

*** Using this Program ***

First, store observed counts in one list.

ex: {8, 14, 20, 20, 24, 41, 20, 29, 10, 9, 5} -> L1

Next, store expected counts in another list.
You may do this in one of three ways:

1. absolute expected frequency
ex: {5.56, 11.11, 16.67, 22.22,27.78, 33.33, 27.78, 22.22,
    16.67, 11.11, 5.56} -> L2

2. relative expected frequency
ex: {.0278, .0556, .0833, .1111, .1389, .1667, .1389, .1111,
    .0833, .0556, .0278} -> L2

3. any values proportional to expected frequency
ex: {1, 2, 3, 4, 5, 6, 5, 4, 3, 2, 1} -> L2

Now, run the CHISQ program and enter the appropriate lists.

ex: OBSERVED: L1
    EXPECTED: L2

The calculator will now display something like:

MEC=5.555555556
df=10
X²=12.361
p=.26162229

explanation of the output:

MEC=minimum expected count.  Often used to determine whether there is sufficient data for a chi-squared test.

df=degrees of freedom

X²=chi-squared test statistic.

p=probability of obtaining a X² value at least as high as the one observed if the expected counts accurately represent the population.