Mirror Sites:
CSUDH - Habermas - UWP - Archives

California State University, Dominguez Hills
University of Wisconsin, Parkside
Soka University Japan - Transcend Art and Peace
Created: October 17, 2003
Latest Update: October 17, 2003
jeannecurran@habermas.org
takata@uwp.edu

Data Entry in SPSS

Site Copyright: Jeanne Curran and Susan R. Takata and Individual Authors, October 2003.
"Fair use" encouraged.

Now that we've had a chance to gather and briefly analyze out data from the recall election in California in 2003, I'd like to take all of us through the steps of playing with that data. I'm going to use release 11.1 of SPSS. The lab uses a different version; the text a different version still. It's important that we learn to switch with ease amongst these versions, since you will encounter many different versions in the field. The SPSS file with our recall data in it is on the site as crs.sav. I think you can access it from the Internet by http://www.csudh.edu/dearhabermas/crs.sav. The dialog box should ask you in what program you want to open the file, and you'll need to open in in SPSS.

The file that I am working with I am calling crsj.sav. You should be able to access that the same way.

Now let's play.

First of all, it's not traditional to call values Qestion 1 and question 2. It's traditional to give them names. Let's try that.

1. Open up SPSS at the data editor by going to the file crs.sav
2. Notice the tabs on the bottom of the screen: Data View and Variable View. Click on variable view to be sure you have the screen that will let you change the names of variables and values.
3. Click in cell (1, Name). The cell will have the rectangle marked by a dark line. Double click, so that the dark line disappears. Now you can change the name. Type the name reclogic in place of Q1.
4. Then click in the cell (1, Label). Double click, so that dark line highlighting cell disappears, and type Recall Logic. and Change Question_1 to Q1.
5. Then click in the cell (1, Missing). Double click, and you will get a dialog box like the one we got by clicking on utilities/variables.
6. In that box click on 0=missing, highlighting it, and then click on remove.
7. Then click in the value box and type 9.
8. Then click in the value label box and type missing.
9. Then click ADD.
10. Notice that the chart is wider than the screen so that after entering the missing value 9 in the missing column, you need to move the scroll bar back to the left to find the name variable again.

What we have just done is to alter the data labels - as we might want to do in a real study. The traditional number for missing data is 9, so I substituted 9 for 0. Some people do use 0, but then you wouldn't get to practice changing it. In McKendrick, you'll find this in Box 3.3 Guidelines for Assigning Missing Values, on p. 59.

Change the rest of the variable list to value names and labels and to missing value = 9.

How come you don't have to change all 85 questionnaires? It's a tricky question. Think about it. You do know the answer. Hint: What is the number 85? How many questionnaires do you have? How many variables do you have? See how easily numbers can confuse you if you don't go back to put them into words. If nobody was confused you're all smarter than I ever imagined, and you'll all have A's in statistics.

Now that you've changed the variable information for SPSS, you need to go back and change the data entry of 0 instead of 9. Remember all your pretty red graphs with that tall column of Missing??? You got those becausse when we switch to the data editor screen, (click on the bottom, just above the task bar. You'll see Data editor and variable view. The data didn't change for 0 to 9 when you changed the variable information. Now you have to go back and change the data. But don't panic. SPSS is wonderful. It does it for you.

Be sure you're in the data editor. You should see all the numbers in the different cells on screen. Then click in the top menu on Transform. You want to transform your 0s to 9s. From the pull down menu choose recode. We're going to recode the 0's to 9's. But this name we want to recode in the same variable. When you click on the same variable the variable dialog box will appear for you to recode. Highlight all the variables, because we want to change all the missing 0's to missing 9's. Then click the arrow, and all the variables will appear in the box to the right of the arrow. Then click old and new variables at the bottom of the dialog box, and the transform dialog box opens. Click the radio button for old value and type in 0; click the radio button for new value and type in 9; then click ADD. Click continue; and then click OK on the next box. Voila. All your 0's have been changed to 9's. Look at the data screen; you'll see they're all 9's now

Once you are sure you have the process for changing variable names and value names and values down, you can skip some of the extra work by just using my file crsj02.sav. I've already changed all the numbers.

Check out the instructions on the Internet: nter-University consortium for Political and Social Research This will give you a sense of areas you might be able to pursue if you like SPSS and statistics. Special Topics Archives at ICPSR.