Chisquare for tables Homogeneity and independence Homogeneity and

Chi-square for tables Homogeneity and independence

Homogeneity and independence • For tables—starting at 2 X 2, there are two kinds of test---with the exact same mechanics— – Independence—is there an influence? • Ho: There is no association between gender and political party (in other words, women and men would join parties at the same rate) or (gender and political party are independent) • Ha: There is an association between gender and political party. (in other words, they are different. ) or (gender and political party are dependent. )

Homogeneity vs. Independence • Homogeneity– sameness • Example: – Ho: Political affiliation for women is the same as it was 60 years ago. – Ha: political affiliation for women is different than it was 60 years ago.

The mechanics GOP DEM 1959 40 60 2009 50 50 Find the expected value first---the expected values assume that Ho is true, in this case that women have the same political party preference as 50 years ago. You do this by Row total X column total / table total

Women and politics Here are the expected cells—you can do this by hand or on the calculator. GOP DEM 1959 45 55 2009 45 55 Then the Chi-square sum will be: (40 -45)^2/45 + (50 -45)^2/45 + (60 -55)^2/55 + (50 -55)^2/55 = a chi square sum of 2. 0202 with degree of freedom of 1 (r-1)(c-1) P-value of. 1552