SPSS Exercises

What can we learn from the General Social Survey (GSS2016) data about the orientations of people who support capital punishment? Is it related to religion? Reflective of attitudes toward race? What about political views? Is it a guy thing? Do attitudes and behavior concerning guns have some relation to support for capital punishment?

1. To answer these questions, we will use some version of each of the following variables in our analysis: PARTYID3, GUNLAW, HELPBLK, DISCAFFF, FUND, OWNGUN, and CAPPUN. Check the wording of each of these questions at the University of Michigan’s GSS website (click on “Browse GSS Variables” and use the mnemonic listing of variables to find those in the list above): www.norc.org/GSS+Website How well does each of these questions meet the guidelines for writing survey questions? What improvements would you suggest?

2. Now generate cross-tabulations to show the relationship between each of these variables, treated as independent variables, and support for capital punishment. A cross-tabulation can be used to display the distribution of responses on the dependent variable for each category of the independent variable. For this purpose, you should substitute several slightly different versions of the variables you just reviewed. From the menu, select Analyze/Descriptive Statistics/Crosstabs:

Rows: CAPPUN
Columns: SEX, PARTYID3, GUNLAW, HELPBLK, DISCAFF, FUND, OWNGUN
Cells: column percentages

(If you have had a statistics course, you will also want to request the chi-square statistic for each of these tables.)

Describe the relationship you have found in the tables, noting the difference in the distribution of the dependent (row) variable—support for capital punishment—between the categories of each of the independent (column) variables.

3. Summarize your findings. What attitudes and characteristics are associated strongly with support for the death penalty?

4. What other hypotheses would you like to test? What else do you think needs to be considered to help you understand the relationships you have identified? For example, should you consider the race of the respondents? Why or why not?

5. Let’s take a minute to learn about recoding variables. If you generate the frequencies for POLVIEWS and for POLVIEWS3, you’ll see how I recoded POLVIEWS3. Why? Because I wanted to use a simple categorization by political party views in the cross-tabulation. You can try to replicate my recoding in SPSS. From the menu, click Transform/Recode/Into different variables. Identify the old variable name and type in the new one. Type in the appropriate sets of old values and the corresponding new values. You may need to check the numerical codes corresponding to the old values with the variable list pull-down menu (the ladder icon with a question mark).