Massachusetts Institute of Technology
Department of Urban Studies and Planning

11.220 Quantitative Reasoning and Statistical Methods for Planning

SPSS

11.220-HOME   |   TEST-OUT   |   BRUSH-UP   |   Brushup-DATA

SPSS > Crosstabs (Chi-square Test)

1. To further explore the relationship between gender and income we can use a cross-tabulation in a two-way table. Go to Analyze->Descriptive Statistics, and click "Crosstabs...."

2. Set "ncws_c" as "Row(s)" and "male" as the "Column(s)." Remember, "ncws_c" is the income variable collapse into three categories. Click on "Statistics."

2. Check "Chi-square" and click on "Continue."

3. Check "Observed" and "Expected." Click "Continue" and then, OK.

4. The crosstab "Counts" option produces the following outcome. We see all of the observed and expected counts in each cell, as well as, column and row totals.

5. Our Null Hypothesis is that the two variables are independent. Using the Chi-Square test statistic we would reject the Null Hypothesis if we got a Chi-Square value greater than the Chi-Square at a particular significance level, and (r-1)(c-1) degrees of freedom (in this case 2 df). At a=.05 and 2 df, we would reject the Null Hypothesis if our calculated C2 is greater than 5.99. The results of this Chi-square test suggest we cannot reject the Null Hypothesis and must conclude that these two variables are independent. In this case, the result may be due to the loss of information that occured when we collapsed the income data into 3 categories, i.e., when we changed it from ratio to ordinal data.

The Linear-by-Linear Association test is a test for trends in a larger-than-2X2 table. It's value is shown to be significant and indicates that income tends to rise with values of "male" (i.e., from 0 to 1).

Created by Myounggu Kang on January 25, 2004. Edited by Rhonda Ryznar on January 20, 2005.