{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Lecture 18_chi-square

Lecture 18_chi-square - 2(Chi-Square Test of Independence 1...

Info iconThis preview shows pages 1–4. Sign up to view the full content.

View Full Document Right Arrow Icon
11/29/2010 1 χ 2 (Chi-Square) Test of Independence 1 2 χ 2 Test Overview Up to this point, we have focused on statistical tests that examined mean differences among two or more groups on the value of an outcome variable We used scores in the sample data to make the inferences about population means Obviously, not all research questions involve continuous outcome measures and/or sample means χ 2 Test Overview The χ 2 is used to examine relationships between two categorical variables when the outcome variable is nominal Recall that nominal variables only allow for qualitative classification--- i.e., they are measured in terms of whether the subjects belong to distinctively different categories 3
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
11/29/2010 2 χ 2 Test Overview More on nominal variables : You do not quantify or rank order the categories of a nominal variable--- no one category is better/worse or higher/lower You cannot say that a given individual has more or less of the quality represented by the variable All you can say is that individuals are different in terms of the variable (e.g., two individuals are of different race or gender) 4 χ 2 Test Overview With nominal (i.e., categorical) outcome variables, you test a null hypothesis that the frequency, or percentage, of people in each category is the same for two or more different groups This is conceptually the same as an independent t-test or an ANOVA, except the dependent variable is nominal--- and thus evaluated in terms of frequencies 5 χ 2 Test Overview You examine observed frequencies in the data to see how many individuals fall in each category The magnitude of the χ 2 test statistic reflects the amount of discrepancy between the observed frequencies you see and the frequencies that you would expect to see if the null hypothesis were true ( i.e. if the variables are not related to one another). 6
Background image of page 2