Data Analysis  
 Comparing Means Across Categories

Now that you have created a new variable, you can use it in any kind of analysis, along with your other variables. I might choose to see if "EQEFFORT," the sense that work is distributed evenly, is related to age or level in school (by running a correlation analysis), but for this example I chose to see if it is related to the way groups organize their work.

Click on "Statistics" (on the top menu bar), then "Compare Means." (You could also click on "ANOVA Models," but "Compare Means" is easier for a beginner to use.)

Again I have several choices. If I want to compare just two categories (Males to Females, for example) I could click on "Independent Samples T Test." Since my Independent Variable has three values, I can use either "One-Way ANOVA" or "Means..." Again I choose "Means" for this example because it is easier for a beginner to use.

I click on ORGWORK, my Independent Variable, and add it to the "Independent List" box.

Notice that I also add GENDER: "Compare Means" can be used with a variable that takes only two values. I then click on my newly created variable, "EQEFFORT," and add it to the Dependent List. If I wanted, I could add other ordinal or ratio-level values to the Dependent List, but I want to keep it simple. I could just click on "OK," but I would like to generate some statistics, so I click on "Options..." and a new window opens.

I make certain that the Cell Displays box is checked to display the Mean, Standard deviation, and Count (number of subjects in each category). I make certain the Variable and value labels are checked. And I check the box for ANOVA table and eta - just for purposes of illustration. (Since my sample is not a random sample and I do not have random assignment to conditions, these statistics will not be very meaningful.)

Now click on "Continue," to return to the first window, and "OK," to run the analysis.

The output from this analysis is shown in the table to the right and below. The top part of the table gives the mean for each group, followed by the standard deviation for each group, the Sum of Squares, and the number of cases in each group. Sum of Squares has to do with the Analysis of Variance (for a detailed explanation, consult a good statistics textbook).

The bottom part of the table gives the output of the Analysis of Variance. If this data set were based on a random sample, or on random assignment to conditions, the results would be significant at p < .05. Given that the data set is not based on a random sample, however, the statistical significance level is not very informative.

I have edited the SPSS outpu to make it fit better in this screen.

Warning: The output from SPSS is still not very easy to read. Do not just copy this into the text of a written report, or onto an overhead transparency! You will need to do considerable work before this table can be presented in a report. I will discuss presentation of results in the next section, Presenting your results.

As noted in the foregoing, SPSS will prepare charts and tables. However, I find the charts and tables prepared by Excel easier to use and format. You can easily copy your results from the SPSS output windows into Excel - or any other Windows based spreadsheet, where you can format them as you like. Just block the material you want to copy and give the "copy" command, either through "Edit" or by hitting "Control-C." You can also copy into your word processing software - but you may have to fix the tabs to get the table back into readable format.

Continue on to Cross Tabs Example



Return to:

Top of this page

SP399 Home Page

David Ritchie's Home Page

Communication Studies
Mainpage
(PSU Home Page)

This page is maintained by David Ritchie, e-mail: cgrd@odin.cc.pdx.edu
Last updated on July 29, 1999.