Chisquares in Excel. Someone. Please. Help.

[KOTKA]

Hey there.
I really hope I have made an effort enough because I am gravely stuck. Deeply, really. Also I can't find anything useful for a Statistic Dummie on the internet, and our class doens't have a book for statistics.

Everything I've done takes time to explain, so for those kind and smart people who can help me the last way to understand I would be oh-so-grateful to.

I have a stats test tomorrow and have covered pretty much everything... except for the last part of how to find the X^2, the statistical parameter of the Chi-square test, and by that, the p-value.

Degrees of freedom is not hard to find (total number of observations subtracted by 2), but the rest... oh my god.

I just need help from someone who knows how to calculate the X^2 in Excel. I have appended the document I've worked with. There are two different tests, and I need help for both of them. The first deals with grades, the other with smoking and cancer. I've made rows & columns of expected values for both and they follow the order of the original values.

Under them, marked as X2, is just... well frankly, I have no idea. I read somewhere that you need to take (expected - original)^2/expected for every cell and plot them up accordingly. They need to be summed up, but how? Is this even correct? Because I have absolutely no idea what to sum up or not.

For those who open my Excel document: Thank you, and forgive me. Something is completely wrong with my Excel since it fails to paste specials. In other words, I tried for one hour to paste in the formulas aswell as the values, so it would be easier for you to follow what I've done, but it didn't work.

I need all the help I can get. I've done all I can and this is my last resort.
My frustration levels are high indeed.

[ocgiraffe]

Hi KOTKA,

This is my first post on these forums but I hope you find some useful information.

Remember there are two types of chi-square analyses the chi-square goodness of fit and the chi-square test of independence.

Now it looks like (from your Excel file) you are running two chi-square test of independence. The degrees of freedom for this test are (# of Rows - 1)*(# of Columns - 1) or (R-1)(C-1). For the first example in your Excel file df = (3-1)(3-1) = 4.

It looks like you calculated the expected values correctly.

So the formula for both chi-square is as you stated:

X2 = SUM[(observed - expected)^2 / Expected]

Which to the right of your observed values you seem to have done correctly for all but one of the cells underneath the expected frequencies. (GRADES 6 and goals)

Now based on the formula above just sum all these up to get your X^2 observed. To do this type =SUM(***) for the *** use your mouse to select all the (observed - expected)^2 / Expected values you calculated

To get the p-value type =CHIDIST(F17,4) where F17 is where I had my calcualted X^2 value from above.

I hope this helps!