Chi-Square Test

# Chi-Square Test - 6) Calculate the degrees of freedom (df)....

This preview shows page 1. Sign up to view the full content.

Steps for a Chi-Square Test for association: 1) Define the Null and Alternative Hypothesis, respectively 0 H and H A . These are in general always “there is no relationship between the 2 variables” and “there is A relationship between the two variables”. When you fill these out, make sure you write the variables in the problem instead of saying “the 2 variables”. 2) (If necessary) Calculate the row, column, and overall totals. 3) Calculate the expected counts. EC = row total * column total overall total . 4) Calculate the partial chi-square values. P.C.S. = 2 (observed count - expected count) expected count . 5) Calculate the chi-square statistic which is the sum of the P.C.S.s.
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 6) Calculate the degrees of freedom (df). df = (# of rows – 1)*(# of columns – 1). Note that you do not include the row of column totals or the column of row totals in the above counts. If this is slightly confusing, it might be best to just look at the table created during step 4. 7) Find the chi-square critical value (from the chart using alpha and df). 8) Draw your conclusion and state it in the context of the problem. Conclusion: Reject H and assert H A if your chi-square statistic is > chi-square critical value. Otherwise, do not reject H ....
View Full Document

## This note was uploaded on 02/06/2012 for the course STAT 225 taught by Professor Martin during the Spring '08 term at Purdue.

Ask a homework question - tutors are online