This guide provides a general walkthrough of SPSS's basic features.

This section and the "Graphics" section provide a quick tutorial for a few common functions in SPSS, primarily to provide the reader with a feel for the SPSS user interface. This is not a comprehensive tutorial, but SPSS itself provides comprehensive tutorials and case studies through it's help menu. SPSS's help menu is more than a quick reference. It provides detailed information on how and when to use SPSS's various menu options. See the "Further Resources" section for more information.

To perform a one sample t-test click "Analyze"→"Compare Means"→"One Sample T-Test" and the following dialog box will appear:

The dialogue allows selection of any scale variable from the box at the left and a test value that represents a hypothetical mean. To find out the probability that the mean of "median_income" is 50000, select the variable from the left-hand box, set the test value to fifty thousand and click "OK." Two tables will appear in the Output Viewer:

The first table gives descriptive statistics about the variable "median_income." The second shows the results of the t_test, including the "t" statistic, the degrees of freedom ("df") the p-value ("Sig."), the difference of the test value from the variable mean, and the upper and lower bounds for a ninety-five percent confidence interval.

In the Data Editor, select "Analyze"→"Compare Means"→"One-Way ANOVA..." to open the dialog box shown below.

To generate the ANOVA statistic the variables chosen cannot have a "Nominal" level of measurement. If you are using the dataset from CSV provided in this guide, the level of measurement will have to be changed to "Ordinal" for the variables "greater_than_thirty_percent_have_bachelors" and "inc_greater_than_ave." This can be easily accomplished in the "Variable View" tab in the data editor (see "Variables in SPSS" under the "SPSS Anatomy" section).

Once the nominal variables have been changed to ordinal, select "high_school" as the dependent variable and "inc_greater_than_ave" as the factor, then click "OK." The following output will appear in the Output Viewer:

To obtain a linear regression select "Analyze"->"Regression"->"Linear" from the menu, calling up the dialog box shown below:

The output of this most basic case produces a summary chart showing R, R-square, and the Standard error of the prediction; an ANOVA chart; and a chart providing statistics on model coefficients:

For Multiple regression, simply add more independent variables in the "Linear Regression" dialogue box. To plot a regression line see the "Legacy Dialogues" section of the "Graphics" tab.

Scholarly Commons