Activity 13.2
Using SPSS for Describing Frequency Distributions

The objective of this activity is to learn how to create frequency distributions with appropriate measures of central tendency and dispersion, using SPSS. To that end, go to SPSS and open the "HOUSE 106depr" data file (download / help downloading). Near the top of the screen, in the status bar area, click on "Analyze." When the drop-down menu appears under "Analyze," click on "Descriptive Statistics." When a box appears to the side, click on "Frequencies."

You are now looking at a new window titled "Frequencies." The small window to the left is the variable list from the data, organized alphabetically. Scroll through the variables until you find the variable "Religion." Click on "Religion," and it will be highlighted. Click on the arrow button between the boxes to move "Religion" into the right-hand box. Then click on the "Statistics" button; notice the choice of measures of central tendency and measures of dispersion available for you to use. Because religion is a nominal level variable, check off only "Mode," and then click on "Continue." Next click on the "Charts" button, which will give you the choice of creating a bar chart or a pie chart. Check off "Pie chart." Another choice in this window is for chart values-frequencies or percentages. Check off "Percentages." Click on "OK" button. Click "OK" again and wait while SPSS calculates a frequency distribution, mode value, and pie chart for the variable religion.

A new window titled "Output-SPSS Viewer" appears; results of your data analysis are in this window. The first table lets you know if there are any missing values for your variable and the value of the mode that you requested. The second table, similar to what is shown in Figure 13.1, shows in the left-hand column the possible values of the variable and, working across from left to right, the frequency, the frequency calculated as a percentage, the valid percentage, and the cumulative percentage. The percentage in the modal category, 56.6%, is the measure of dispersion.

At the top of screen, click on "Analyze," and repeat the preceding process for the variable "ptyunity," which stands for party unity and is the percentage of time that the representative voted with a majority of his or her party. But before you click on the "OK" button this time, be sure to move the "Religion" variable back into the left-hand box, or SPSS will do a frequency distribution for that variable again. You can do a frequency for as many variables in a data set as you want at one time by moving them all into the right-hand box at the same time. SPSS will do a frequency simultaneously for all the variables listed in the right-hand box, but remember that variables operationalized at different levels will require you to choose different measures of central tendency and dispersion. Because party unity is measured at the interval level, you will choose different options for central tendency and dispersion measures.

Click on the "Statistics" button and check off mean, median, mode for measures of central tendency and range, variance, and standard deviation for measures of dispersion. Click on "Continue." Click on the "Charts" button, and check off "Bar chart" and "Frequencies." Click on the "OK" button, and wait while SPSS calculates a frequency distribution, measures of central tendency and dispersion, and a bar chart for the variable of party unity.

To print only partial results of your SPSS output, such as only one table, right-click on that table and a box will appear around it. For example, when doing a frequency distribution of a variable with many values that produces a very large table that might take up several pages, you may want to print only the table reporting your measures of central tendency and dispersion. Click on "File" near the top of the window, and when the drop-down menu appears, click on "Print." A window titled "Print" appears. Click on "Print preview" to see if only the table you have designated appears. Click on the "OK" button to print.

Write a brief description of the frequency distributions of religion and party unity for U.S. House members, making use of all the measures of central tendency and dispersion for each variable.