This Pass or Prepared? is brought to you, courtesy of the Social Science Research Institute Council, from their Workshop on Wednesday, June 14, 2000. Berkeley has put up Survey Documentation and Analysis (SDA) that will let you run frequencies and crosstabs from your home computer, using real survey data banks.

Let's try it.

• First, link to Berkeley's SDA Site

• Then, with the Berkeley SDA site on screen, click the Start button in your taskbar. Click on Netscape Navigator, or whatever browser you are using. You should now have two buttons for your browser on your taskbar. You can move from one to the other just by clicking on the taskbar button. This will enable you to switch back and forth quickly between two different (or three or four . . . different) screens. We're going to want to switch from Berkeley's SDA site to this exercise. So, when you have the Berkeley SDA site on one of the screens, put up http://www.csudh.edu/dearhabermas/pssda01.htm on the other. Switch back and forth until you get the hang of it.

• Now, with the Berkeley SDA site on one screen and the exercise on the other, let's try the program. With the Berkeley site on your screen, click on the Calpolls link in the left hand frame. Notice that the left most frame does not change, but that the right hand frame is now split into two separate screens. The screen on the left gives an overview of data banks that can be used with SDA. The screen on the right gives a search tool for the California Field Polls data base at the University of California, San Diego.

• Since the UC Data for SDA use are limited for the time being to the sets described in the left hand frame, let's choose a set that we know is up. Scroll down about four-fifths of the way to 94-05. Link on 94-05. Notice that the right hand frame now gives you the codebook for that data set.

• Click on Headings for Groups of Variables, in the right hand frame. Scroll down to the opinion of death penalty variable, about four fifths of way down, under Opinion of O.J. Simpson issues. Note that the variable is labelled v180.

• Continue to scroll down in the right hand frame to Demographics. Notice the variable v193, based on answers to the question "Do you consider yourself a strong or not very strong conservative or liberal?

• Consider the hypothesis that those who are most conservative favor the death penalty, while those who are most liberal oppose the death penalty. The theory behind such a hypothesis might be that the death penalty is the ultimate retribution, and that the liberal philosophy leans more heavily towards forgiveness. Let's run a crosstab on that hypothesis. The way we have stated the theory, if we knew the political philosophy of a subject, we believe that we will be better able to predict that subject's position on the death penalty. That is, we conceptualized political philosophy as the independent variable from which we will try to predict attitude towards death penalty, the dependent variable. Dependent, because in our theory the attitude to the death penalty depends on the person's political philosophy. Let's see if we are right.

• In the left hand frame enter the independent variable label, v193, strength of political identification as conservative or democratic, in Vertical. Enter the dependent variable label, v180, attitude towards death penalty, in Horizontal. Scroll down to Percentaging and click on Vertical. Then click on Question text; this prints the question itself on your output to help you with interpretation. Now click on the gray burron that says Run the Table.

• In the left hand frame click on the run crosstabulation radio button. Then press the gray Start button.

• The table appears in the left hand frame. You will need to scroll around to see it. But you can print a hard copy by clicking on your browser printer icon.

• Let's read the table.

  Click on any question number to see jeanne's answer to that question.

  1. What percentage of not very strong liberals would do away with the death penalty in 1994 in this California poll?

     One Plausible Answer

     30.3 percent of not very strong liberals would do away with the death penalty.

  2. How many people does that represent?

     One Plausible Answer

     30.3 % represents 23 people.

  3. Why are the data in questions 1 and 2 confusing.

     One Plausible Answer

     Because we usually give percentages to simplify data, to take very large numbers and make them more understandable. In this case 30% sounds more impressive than 23 people. Always be sure to check the actual number of respondents. Five dentists recommending someone's favorite toothpaste may be 60% of the sample. This will lead to problems of interpretation. It's hard to generalize from five dentists. We'll talk more about this in inferential statistics. That is, what may we infer about toothpaste generally from what our sample of dentists tells us.

  4. Is there a fairly equal distribution of the sample over the liberal/conservative spectrum? Or, put another way, are liberals and conservatives pretty much balanced in California?

     One Plausible Answer

     A frequency table would answer that quickly. Glance over at the right hand frame. v193 would provide us with this answer. The frequency distribution is given along with the question. 35% are strong conservatives as opposed to 20% strong liberals.

  5. What does that say about the frequency distribution of liberals and conservatives?

     One Plausible Answer

     Well, one thing we could do is add the strong and not very strong conservatives together, and the strong and not very stron liberals together. Then we could compare theconservative/liberal frequency distribution and run that crosstabulation with the attitude toward the death penalty. This is called recoding. go back to the Berkeley SDA analysis screen by clicking on the screen and then pressing the Back button on your browser. go back to the original request to run the data.

     In the vertical dialog box, Open r: (1-2; 3-4). Then rerun the table.

     We're gonna transfer this to a separate exercise.

     . . .

  ---

  ---

  Figurine by Rudiger Appel. Notice that you can see three effects in the animation. Either the Variation on the Kandinsky figurine appears to turn in a clockwise direction, or in a counterclockwise direction, or it appears to open and close. Can you see all three effects? Try. Fascinated? Link to Appel's site and then link to the background he provides. Scroll down until you find a link to background.

Copright: Jeanne Curran and Susan R. Takata, June 2000. "Fair Use" encouraged.