1. Introduction to SPSS July 28, 2008 2:00-4:00 pm 112A Stright Hall
By Paul Rossman
Applied Research Lab
G10 Suites On Grant Upper

2. Abstract
This workshop is designed for users who wish to enter and analyze their survey data in SPSS. After this workshop, you should be able to:
1) Create an SPSS workbook for your survey. 2) Code your survey responses in SPSS. 3) Verify your survey data. 4) Create graphs and charts of your data. 5) Generate descriptive statistics from your data.

3. Supporting Materials (See Appendix A)
The Attitudes Toward Research Scale (ATRS) that measures attitudes toward research.
developed by Elena C. Papanastasiou.
Respondents select the response that best describes how strongly they agree or disagree with each of 31 statements covering aspects of educational research.
The items form five subscales: (a) Usefulness of research for the Profession, (b) Research Anxiety, (c) Positive Attitudes Toward Research, (d) Relevance of Research to Life, (e) Perceived Difficulty of Research.

4. During Summer 2008, the ARL wrote a StudentVoice survey that was randomly sent to 600 IUP graduate students.
Asked for demographic information (age, race, gender, program, degree, etc).
Asked the students to complete the ATRS.
56 students responded.
See Appendix A for a copy of the survey

5. A. Introduction
SPSS (Statistical Package for the Social Sciences) is one of the most widely available and powerful statistical software packages.
What is an SPSS dataset?
What are the benefits of learning SPSS?

6. B. Finding SPSS In Windows: Click on the “Start” button.
Scroll up to “Programs”
Select “ SPSS for Windows”.
Select “ SPSS 15.0 for Windows”.

7. C. Verifying your data In SPSS:Select the “Data” pull-down menu.
Select the “Define Variable Properties” option.

8. Move the variables you want to check from the box on the left to the box on the right, by highlighting them and clicking on the arrow between the boxes.
Click on the “Continue” button.

9. 5. You will see a window like this
Notes
See highlighted row: It says that 24 respondents moderately
agreed with the statement. “Moderately agree” is coded 4 in SPSS.
b) When analyzing survey data, “Unable to judge” is considered a missing value. Mark it as missing. See next slide.

10. Copy missing value to all of remaining items (Q17-Q46)
Click on the “To Other Variables” button. Then highlight variables Q17-Q46, and click on the “Copy” button.
When SPSS sends you back to the previous menu, click on the “OK” button.

11. D. Looking at the “Other” responses
It is always a good idea to see what responses users gave for “Other”. See Q6 for example.
Click on the “Analyze” pull-down menu.
Select the “Descriptive Statistics”, then the “Frequencies…” options.
Move “Q06 degree other” to the “Variable(s):” box.
Click on the “OK” button.

12. Need to change the codes for Q06 degree for these 10 respondents.

13. Easiest Way--Sort cases by Q06 degree other.
From the “Data” pull-down menu, select the “Sort Cases…” option.
Choose the “Q06 degree other” variable.
Click on “OK” button.

14. At the bottom of the dataset in “Data View”, you see:

15. E. Describing your Variable Attributes
At the bottom of the screen, click “Variable View”.
Let’s modify the attributes for Q1 (row 2).
Change the variable name from “Q1” to “Q01”
Change the type to “Numeric”.
Change the decimals from 2 to 0.
Enter a relevant label. A good one would be “Q01 gender”.
Add value labels (see next slide).
Change the Measure from “scale” to “nominal”.

16. Add Value Labels (step 7 on previous slide)
Gender has two values, so we will tell SPSS that. Under values click on “None” and a box with three ellipses should appear. Click on that box and a window called “Value labels” should appear.
In our survey, a value of 1 indicates female and a value of 2 indicates male. So we will tell SPSS that. In the box next to “Value:”, type in 1. In the box next to “Label:”, type in female. The click “Add”. Then, in a similar manner, label the male respondents. When you are finished click “OK”.

17. F. Recoding Q04 into new binary variable Q04_N
Because there are very few respondents who are not white/caucasian,
Select the “Transform” pull-down menu.
Select the “Recode into Different Variables” option.

18. Move the variable you want to recode from the box on the left to the box on the right, by highlighting it and clicking on the arrow between the boxes.
Type in the new variable name and label.
Click on the “Change” button.
Click on the “ Old and New Values… ” button.

19. Put in the coded value (3) for “Caucasian” in the “Old Value” box and 1 in the “New Value” box. Then click on the “Add” button.
Select the “All other values” option. Then type in 0 in the “New Value” box and click on the “Add” button.
Click on the “Continue” button.

20. G. Descriptive Statistics Summary statistics for a scale variable
Select “Analyze” pull-down menu.
Select “Descriptive Statistics” option.
Select “Descriptives…” option.

21. Let’s do min, max, range, mean, and standard deviation for “Q02 age”
Let’s do min, max, range, mean, and standard deviation for “Q02 age”. Move Q02 age to the “Variable(s):” box.
Click on the “Options…” button.
Select the desired statistics: min, max, range, mean, and standard deviation.
Click on the “Continue” button.
Click on the “OK” button.

22. SPSS Means Output

23. Histogram of the Respondents Ages (Histograms are used for scale variables)
Select the “Graphs” pull-down menu.
Select the “Legacy Dialogs” option.
Select the “Histogram” option.

24. Move the variable “Q02 age” to the “Variable:” box.
Click on the “OK” button.

25. SPSS Histogram Output

26. H. Descriptive Statistics Summary statistics for a scale variable
Select “Analyze” pull-down menu.
Select “Compare Means” option.
Select “Means…” option.

27. Move the scale variable (Q02 age) to the “Dependent List:” box.
Move the “by” variable (Q02 gender) to the “Independent List:” box.
Click on the “Options…” button.
Move the desired statistics over to the “Cell Statistics:” box.
Click on the “Continue” button”.
Click on the “OK” button”

28. SPSS Means Output

29. Histogram of the Respondents Ages By Gender
Select the “Graphs” pull-down menu.
Select the “Legacy Dialogs” option.
Select the “Histogram…” option.

30. Move the variable “Q02 age” to the “Variable:” box.
Move Q01 gender to the “Rows:” box.
Click on the “OK” button.

31. Side-By-Side Box Plots of the Respondents Ages By Gender
Select the “Graphs” pull-down menu.
Select the “Legacy Dialogs” option.
Select the “BoxPlot…” option.
Select the “Simple” Option”
Select the “Summaries for Groups of Variables” option.
Click on the “Define” button.

32. 7. Move the “Q02 age” variable to the “Variable” box.
8. Move the “Q01 gender” variable to the “Category Axis” box.
9. Click on the “OK” button.

33. SPSS BoxPlot Output

34. I. Cross Tabulation Select the “Analyze” pull-down menu.
Select the “Descriptive Statistics” option, then select the “Crosstabs…” option.
Move one categorical variable (Q01 gender) into the “Rows” box and the other categorical variable (Q07 campus) into the “Columns” box.
Select the “Cells…” button.

35. 6. Under Percentages, select “Row”, “Column” and “Total”.
5. Under Counts, select “Observed” and “Expected.”
6. Under Percentages, select “Row”, “Column” and “Total”.
7. Then click on the “Continue” button.
8. Finally, click on the “OK” button.

36. SPSS Crosstab output
Interpretation of the numbers in the highlighted cell:
12 of the 56 respondents (21.4%) are male and study primarily at the Indiana Campus.
If there were no relationship between Gender and Campus, you would expect to see 10.8 respondents who are male and study primarily at the Indiana Campus.
12 of the 14 male respondents (85.7%) study primarily at the Indiana Campus.
12 of the 43 respondents (27.9%) from the Indiana campus are male.

37. Can you produce this bar graph?

38. Can you produce this bar graph? Hint: you need a new variable.

39. Can you produce this matrix scatter plot?

40. J. Resources Applied Research Lab: http://www.iup.edu/arl/
Which statistical test should I use?
4. Student version of SPSS is available at the IT Support Center in the Suites on Grant Lower.