SIMEON Is a Trademark of the Esys Corporation

School of PsychologySPSS v.11 for Windows XPSPSSv13-XP Analysis CPD.doc

Descriptive Statistics

The first thing that you want to do once you have entered all your data into a spreadsheet is to summarise it. The statistics that summarise data are called 'descriptive statistics'.

Calculating descriptive statistics

Click on Analyze (sic) in the main menu bar, then select Descriptive statistics and click on Frequencies.

Select those variables you want to report, but now click on Statistics.

Check the boxes relating to those statistics you want to report. Click on Continue.

Click on Charts. Select Bar chart, click on Continue and then on OK.

Plotting a histogram

Click on Graphs… …Histogram

Select the relevant variable(s) and click on the  button to move them across to the ‘Variable(s)’ box.

Click on OK

Try experimenting with some of the other options that appear when you click on the Graphs option in the top menu bar. Which format is most suitable for the data that you want to represent?

Inferential statistics

Choosing the right test depends upon the type of data that you want to test. 'Between subjects' or 'non-repeated measures' means that the two groups that you wish to compare contain entirely different participants. 'Within subjects' or 'repeated measures' means that two 'scores' are taken from each respondent, and these are then compared (eg: 'before' vs 'after' measures).

‘Parametric’ means that the data that you are analysing is distributed normally, or is close to a normal distribution. If you consider that the variable is normally distributed in the general population, it does not matter whether or not the same is true for the sample that you are analysing. You can still use parametric tests in these circumstances.

T-test: Comparing two groups' scores on a continuous variable

There are two types of t-test; one for two unrelated groups of data ('between subjects', eg: male vs female) and one for related groups of data ('within subjects', eg: scores on a variable for the same participants 'before' vs 'after' some event).

Related (Within subjects) test:

Click on Analyze in the main menu bar, then select Compare Means, and click on Paired Samples T-Test.

Select the two variables that represent the two scores that you want to compare (note that you must enter these in pairs - again, think of the logic behind this).

Click on OK.

Unrelated (Between subjects) test:

Click on Analyze in the main menu bar, then select Compare Means, and click on Independent Samples T-Test.

Select the variable that you want to test for differences, and put it in the Test Variable(s) box.

Select the variable that distinguishes between the two groups that you want to test (eg: 'gender') and put it in the Grouping Variable box.

Click on Define Groups, and type in the numbers that you have used to identify the two groups (for example, if you have used a "1" for males and a "2" for females, type a "1" in the 'Group 1' box and a "2" in the 'Group 2' box). Then click on Continue.

Click on OK.

Correlation: comparing two continuous variables

Click on Analyze… …Correlate… …Bivariate

Move the two variables for comparison into the Variables box.

Ensure Pearson's is selected for Ratio (continuous) variables or Spearman's for Ordinal (ordered category or ranked) variables.

Click on OK.

Chi-square: Categorical (frequency) data - one variable

When you have an ordered or unordered category variable that records, for example, the number of times that an event occurred or respondents’ favourite colour, you should use the chi-square test.

Click on Analyze in the top menu, then select Nonparametric Tests, and click on Chi-Square.

Select the variable(s) that differentiates the categories that you want to compare, moving them into the Test Variable List box.

Click on OK.

Cross-tabs: Categorical (frequency) data - two or more variables

Click on Analyze… …Descriptive Statistics, and click on Crosstabs.

Select the two variables that you want to compare, put one in the Row box, and one in the Column box.

Click on Statistics, and check the Chi-Square box. Click on Continue.

Click on Cells and select Expected in the Counts section (you can also select the row/column/total Percentages option if you want)

Click on OK.

The Pearson’s Chi-Square statistic tells you if there are significant differences in the number of observations falling into each of the cells of the crosstabs table. Comparing the ‘observed’ frequencies with the ‘expected’ frequencies will tell you where those significant differences lie.

Wilcoxon: Repeated measures (two categories)

Click on Analyze… …Nonparametric tests… …2 Related samples

Select the two variables that you want to test for differences and move them into the Test Pair[s] List box (for example: the columns containing before and after measures).

Ensure that the Wilcoxon option is checked in the Test Type section.

Click on OK.

Mann-Witney: Non-repeated measures (two categories)

Click on Analyze… …Nonparametric tests… …2 Independent samples

Select the variable that you want to test for differences and move it into the Test Variable List box.

Select the variable that distinguishes between the two category groups and move it into the Grouping Variable box.

Now click on Define Groups and type in the coding for each group (for example: if you want to compare male vs female, and male=1 while female=2, then type a '1' in the first box and a '2' in the second box. Click on Continue.

Ensure that the Mann-Witney option is checked in the Test Type section.

Click on OK.

Friedman: Repeated measures (three or more categories)

Click on Analyze… …Nonparametric tests… …K Related samples

Select the variable that you want to test for differences and move it into the Test Variable List box.

Select the three or more variables that you want to test for differences and move them into the Test Variables box (for example: the columns containing before, during and after measures).

Ensure that the Friedman option is checked in the Test Type section.

Click on OK.

Kruskal-Wallis: Non-repeated measures (three or more categories)

Click on Analyze… …Nonparametric tests… …K Independent samples

Select the variable that you want to test for differences and move it into the Test Variable List box.

Select the variable that distinguishes between the category groups (for example: First, Second and Third-Year students) and move it into the Grouping Variable box.

Now click on Define Groups and type in the coding range for the groups that you want to compare (for example: if you want to compare First, Second and Third-Year students, and First=1, Second=2 and Third=3, then type a '1' in the 'Minimum' box and a '3' in the 'Maximum' box. Click on Continue.

Ensure that the Kruskal-Wallis option is checked in the Test Type section.

Click on OK.

If you have any problems interpreting the output from any of the above procedures, please ask one of the demonstrators for help.

You may also want to take a look at the flow diagram on the web at:

It should help you decide which test you need for a particular set of variables, and will give you the relevant SPSS instructions.