SPSS Intro and Crosstabs
LSP 121
Activity 2b – 20 points
Crosstabs and SPSS
1) Let’s try another example using SPSS. In this example, we will enter our own data and then perform a crosstab, or cross-tabulation. A crosstab allows us to make comparisons of survey data across classifications. In surveys, the classifications are usually age, race, gender, party affiliation, etc.
For example, assume that a number of college students were asked how many schools they visited before selecting a college. We also recorded each student’s sex and assigned a simple ID. Thus, we might have data that looks like the following:
Gender
ID 1=Female, 2=Male Schools visited
11111 1 3
11112 2 4
11113 2 7
11114 1 5
11115 1 2
11116 2 1
11117 2 6
11118 1 4
11119 1 2
We would like to generate for each sex a count of how many visited 1 school, how many visited 2, etc. Thus, we would like to generate something that looks like the following:
Schools visitedGender / 1 / 2 / 3 / 4 / 5 / 6 / 7 / Grand Total
Female / 2 / 1 / 1 / 1 / 5
Male / 1 / 1 / 1 / 1 / 4
Grand Total / 1 / 2 / 1 / 2 / 1 / 1 / 1 / 9
Open SPSS and enter variables: ID, gender, schools (all 3 are numeric, decimals = 0). Type in the data from the survey in the data view mode. Add the values in the values column: (Gender: 1 = female, 2 = male). Note: you enter the value, the label and then click ‘Add’ to add what you just entered.
Then pull down the menu Analyze and click on Descriptive Statistics, then Crosstabs. What variable do you want in the row? The column? When ready, click OK to perform the crosstab. (Hint: ID should not be in either the row or the column. You are not interested in analyzing the IDs – you are interested in analyzing the visits and the corresponding sex.)
Paste the cross-tab output into your Word document. Make sure the labels in Word document make sense. It is possible to edit the text in the Word document if necessary.
2) Look at a larger example. Do a crosstab on the Excel file MCIC-SMOKE.xls. (File is in the ‘older data’ group.) Look at the data in Excel before you import this into SPSS. Note that tab 1 contains the codes for the data for ’93 and ’97. This Excel file consists of the 1993 and 1997responses for the question "Do you smoke now?"(SMOKE) along with a set of demographic variables: age (AG), race/ethnicity (RACETH), income level (INC4GP), religious affiliation (REL1), educational level (RESPED), and gender (SEX01). Note that the 1993 and 1997 data are on separate worksheets. There is also a code worksheet that tells you what the codes mean. Familiarize yourself with the codes for SMOKE and SEX01. Open this file in SPSS; be sure to select the 1997 sheet. In ‘Variable View’ tab, change the Values (column 6) to the appropriate equivalents, e.g., for SEX01, 1=Male, 2=Female and for SMOKE, 1=Yes, 2=NO. It is a common choice in a survey to allow the user to select ‘Refused’ which indicates ‘refused to give an answer’. Not applicable here.
We would like to find out what percentage of men smoked and what percentage of women smoked in the sample in 1997. The results should produce a diagram that looks something like (numbers will differ, color is from Excel view):
Men / Women / TotalSmoke / 20 / 25 / 45
Not Smoke / 80 / 125 / 205
Total / 100 / 150 / 250
Notice how the numbers add up going down and going across. Using SPSS, create a crosstab and paste the crosstab output into your Word file into question 8. Explain in sentences how to interpret the cross-tabulation and how you would determine the percentages.
3) Would it be interesting to compare the smoking rate to another variable? For example, which race/ethnicity smokes or does not smoke? Select one of the other variables and produce a crosstab comparing that variable to smoking. Select the output of this cross-tabulation in SPSS and paste it into your Word Document. Write a short paragraph describing what variables you just compared and why and include this discussion with question 1 and 2.
Include names of group members in Word document. Submit Activity 2b as a Word document to DropBox for activity 2b.
1