SPSS: Chi-square test for Independence

The data example:  Are the same proportions of blonde, brunette, and red hair found in both the male and female student populations? 

STEP 1:  Define variable name, label, and values

Open SPSS and choose Type in data. Start by naming the first variable of interest (the row variable). Click the Variable View sheet tab.  Type the variable name in the Name column. Specify variable type, width, and number of decimals. You have the option of typing a more descriptive variable name (255 characters max) in the Label column.  Click on the values cell and then the gray box in the right corner of the cell.  A Values dialog box opens.  Label the categories of the variable (e.g., for "Gender", "1.00 = Female", "2.00 = Male").  Click the Add button after defining each value.  Repeat the above steps for the second variable (the column variable).

STEP 2: Input the data into SPSS

Make sure to click back into Data View.  The data are frequencies, not scores, so you will be using a different numeral to specify a subject in each category. Each subject's response can be classified along two variables (gender and hair color).  Essentially, you will be entering the data twice, once for the row variable and again for the column variable. You will be using two SPSS variable columns for the data. Enter a 1 for all the elements in the first row (females) and a 2 for each element in the second row (males). Next, in the second SPSS column enter the data for the columns (hair color), working across the first row, then across the second row). As you’re working, make sure you verify you’re working with the correct information on the first variable.  The total number in the two columns should match.

STEP 3: Select the statistical procedures

Go to the Analyze menu and select Descriptive Statistics and then select Crosstabs.

STEP 4: Select the variables

Inside the crosstabs dialog box, select the variables to analyze (for example, row variable is Gender and column variable is Hair) by using the arrow keys to move the variable names into the right box. Next, click on the Statistics box at the bottom corner and select Chi-square. Next, click on the Cells box at the bottom corner and select both the observed and expected counts options; this will print out both the observed and expected frequencies in your table. Click OK. The observed and expected frequency tables, along with the chi-square statistic will be displayed in your output window.

STEP 5: Look up the critical value of chi-square or look at the significance level reported on the output and make a conclusion regarding Ho. Or use the significance level reported in the output to make a decision regarding Ho.