Chi sqaure test - problem with DF and interpretation

Hello everyone,

I wish I could not bother you but I have some difficulties with my MA thesis research. I would strongly appreciate if someone could help me.
So, here is the deal. I have two sets of test scores. One set is the observed one and the second one is the expected one. The writing test scores are categorized with respect to four categories (such as Vocabulary, Grammar, Content and Organisation). There are 22 two tests. So, the table (of one set - there are two sets) has 4 columns (for categories) and 22 row (for the scores, of 22 people). The range of points one can get in one category is from 0 to 10.

My instructor told me to compute the CHI SQUARE TEST. However, as a closed-minded humanist I failed (at least, I think so). I tried counting it in Excel – where the score was 0,491316 and in Statistica (which I think I do not know how to use) the score was 64,08548. If one of those scores would be correct I still do not know what to do with it.

It seems to me that the Statistica score is more probable, however, I do not know what will be the DF value for me to find the probability rate in the table.

I will appreciate and be grateful for every suggestion and feel free to mock me if my question sounds banal. Thanks!


Ambassador to the humans
What hypothesis do you want to test?

Are the "expected scores" the same for all 22 people or are they different for everybody?
Thank you for your reply.

After tons of reading and experimenting I managed to calculate the Chi Square (which is 29,85987711). I established that the DF (Degrees of Freedom) is equal 21 (since there are 22 writing samples - but you have to subtract 1). To clarify something, Observed scores are those which I am trying to prove to be reliable, correct, good or whatever, and the expected are those, which are reliable, correct, good.

The expected scores differ. The minimum is 0 and the maximum is 10. However, in reality the lowest given is 3 whereas the highest is 9.

I do not have a clearly stated hypothesis, however, the point is to see whether the scores (observed ones) are reliable (in general, and with respect to the expected ones). So, since I calculated the CHI Square and confronted the result with the table of Values of the Chi-squared distribution table it appears that the differences between the two set of scores do not have the statistical significance, therefore are marginal and not relevant. Please correct me if I am wrong. If you have any other conclusions in mind that are, or might be, observed from those calculations I would strongly appreciate if you could share them with me (whenever you find some spare time). Thank you in advance.