Quick Reference for Basic SPSS Procedures
After you have put your data in a data file using the Data Editor, you can use SPSS to compute all sorts of statistics. This handout provides a brief guide to some common statistics. If you need more information about these SPSS statements or other statistics, you can use the HELP command in SPSS.
COMPUTING DESCRIPTIVE STATISTICS (means, standard deviations, variance, etc.)
To obtain descriptive statistics, from the menus choose:
 |
Analyze |
|
|
Descriptive Statistics |
|
|
Descriptives |
Your variables will appear on the list on the left. Select one or more variables for which you want statistics by clicking on the variable and then clicking on the arrow. When you have chosen all variables, click on OK.
If you want statistics other than the standard (mean, standard deviation, min, max), click on Options, select them, go back to the Descriptives menu (click on Continue) and then click on OK.
COMPUTING FREQUENCIES (a tally of values for selected variables)
To obtain frequencies, from the menus choose:
|
|
Analyze |
|
|
Descriptive Statistics |
|
|
Frequencies |
Your variables will appear on the list on the left. Select one or more variables for which you want frequencies by clicking on the variable and then clicking on the arrow. When you have chosen all variables, click on OK.
To obtain additional statistics for these variables, in addition to frequencies, click on Statistics and select the statistics you want computed, then click on Continue to go back to the Frequencies menu. Click on OK.
To obtain a histogram or bar chart, click on Charts, select the type of chart you want, then click on Continue to get back to the Frequencies menu. Click on OK to run procedure.
COMPUTING A CROSS TABULATION (frequency of values for
one variable with another variable)
To obtain a cross tabulation, from the menus choose:
|
|
Analyze |
|
|
Descriptive Statistics |
|
|
Crosstabs |
Your variables will appear on the list on the left. Select one variable and click on the arrow next to the Row(s) list box. This moves that variable to the Row(s) list. Then select a variable and click on the arrow next to the Column(s) list. This moves this variable to the Column list. Now click on OK to run the crosstab.
To obtain other statistics for these variables (such as Chi-Square or correlations), click on Statistics, select the statistics you want computed, then click on Continue to go back to the Crosstabs menu. Then click on OK to obtain the statistics.
To obtain percentages for the cells, rows, and/or columns, click on Cells (on the Crosstabs menu), select the options you want, and then click on Continue to go back to the Crosstabs menu. Click on OK to run the procedure.
COMPUTING CORRELATIONS (Pearson product moment correlations)
To obtain "r" choose:
|
|
Analyze |
|
|
Correlate |
|
|
Bivariate |
Your variables will appear on the list at the left. Select variables for analysis by highlighting the variable and then click on the arrow. You have to select at least two variables to compute a correlation. The default (standard) setting displays a two-tailed Pearson product -moment correlation. This is usually what you want. To obtain the correlations, click on OK.
COMPUTING A LINEAR REGRESSION ANALYSIS
To obtain a regression analysis, from the menus choose:
|
|
Analyze |
|
|
Regression |
|
|
Linear |
Your variables will appear on the list at the left. Select the "dependent" variable (criterion/outcome variable) by highlighting the variable and clicking on the arrow. Now select one or more "independent" variables (predictor variables) by highlighting the variable and then clicking on the other arrow. If you want the default forced entry procedure, just click on OK to get the analysis. If you want another type of analysis (e.g., stepwise, backward, forward), change the Method using the arrow key next to the method box to select the procedure, then click on OK.
To obtain additional statistics for your variables, click on the Statistics box, then on Descriptives, and select means, standard deviations, and/or correlations. Click on Continue until you are back at the Linear Regression menu. Click on OK to run the procedure.
You can also obtain a plot of your data by clicking on Plots and choosing the desired options.
OBTAINING A SIMPLE SCATTERPLOT (graph of two variables)
To obtain a scatterplot, from the menus choose:
|
|
Graphs |
|
|
Scatter |
This opens the scatterplot menu, highlight the Simple picture and click on Define. Your variables will be listed on the left. Highlight the variable you want on the Y-axis and click on the Y axis arrow. Then, highlight the variable for the X-axis and click on the X axis arrow. To get a simple scatterplot in default form, click on OK.
OBTAINING MEANS FOR SUBGROUPS OF THE DATA (e.g., means for GPA for males versus females)
To obtain means for subgroups in the data set, choose from the menus:
|
|
Analyze |
|
|
Compare Means |
|
|
Means |
This opens the Means menu. Select your criterion/outcome variable (GPA in my example) by highlighting it in your variable list and clicking on the arrow. Now select your predictor variable (sex in my example) and click on the bottom arrow. Click on OK. This will give you the means, standard deviations, and number of cases for males and females for GPA.
SELECTING ONLY CERTAIN CASES TO ANALYZE (such as only the data for the females)
To select a subset of cases for analysis, from the menus choose:
|
|
Data |
|
|
Select Cases |
This opens the Select Cases menu. Choose the
Filtered alternative for the treatment of the unselected cases (e.g., males in
my example). If you choose Deleted the other cases will be deleted from the data
file. You usually don’t want this to happen!
Now you have to select your cases based on some conditional statement. So,
select the If condition is satisfied statement by clicking on
IF. In the box, type the statement that defines the
cases you want to select for analysis (e.g., sex=2) or use the variable list,
calculator, etc. to paste the information in the box. Then, click on
Continue and OK. After
choosing your subset of the data, all future analyses will only use that sample.
To include all data, go back into the Select Cases
menu and select All Cases, then click on
OK.
COMPUTING AN INDEPENDENT GROUPS T-TEST
To obtain this t-test, from the menus choose:
|
|
Analyze |
|
|
Compare Means |
|
|
Independent-Samples T Test |
This opens the Independent-Samples T-Test menu. Choose one or more dependent variables (called test variables in the menu) by highlighting them on the variables list and then clicking on the arrow. Then choose the independent variable (called group variable) and click on the other arrow. You must define two groups for the independent (grouping variable). Highlight the grouping variable and click on Define Groups. Most of the time you will want the default, Use Specified Variables. Simply enter the value that corresponds to group 1 (e.g., For the sex variable, 1 corresponds to males) and then the value that corresponds to group 2 (e.g., 2 corresponds to females). Click on Continue and then on OK to run the procedure.
COMPUTING A PAIRED (OR CORRELATED) SAMPLES T TEST
To obtain this type of t-test, choose from the menu:
|
|
Analyze |
|
|
Compare Means |
|
|
Paired-Samples T Test |
This opens the Paired-Samples T Test menu. Click on the first variable (e.g., Pretest). Click on the second variable that you are comparing to the first variable (e.g., Posttest). They will appear in the Current Selections box. Click on the arrow to move the variables to the Paired Variables list. Click on OK to run the procedure.
COMPUTING A ONE-WAY ANALYSIS OF VARIANCE (1 independent variable with 3 or more conditions)
To obtain a one-way ANOVA, from the menu choose:
|
|
Analyze |
|
|
Compare Means |
|
|
One-Way ANOVA |
This opens the One-Way ANOVA menu. Select your dependent variable and click on the first arrow. Then select your independent variable and click on the other arrow. Then click on OK to run the procedure.
To obtain means and standard deviations for your groups, click on Options, select Descriptive, and click on Continue. To get Post-hoc Comparisons or Contrasts, click on the option you want from the One-Way ANOVA menu. Then click on OK to obtain the ANOVA and other statistics.
COMPUTING A SIMPLE FACTORIAL ANALYSIS OF VARIANCE (one-way or two way anova)
To obtain an analysis of variance when you have one or more than one independent variables, choose from the menus:
|
|
Analyze |
|
|
General Linear Model |
|
|
Univariate |
This opens the Univariate menu. Note that "univariate" refers to the number of dependent or outcome variables (i.e., one), not the number of independent or predictor variables. Select your dependent variable and click on the first arrow. Then select your independent variable(s) (Fixed Factors) and click on the second arrow. Click on OK to run the test.
To obtain means and standard deviations for your dependent variables by the levels of your independent variables (Hint: you will always want to do this!!), before running the procedure, click on Options. Click on the box for Descriptive statistics. Click on Continue and then OK to run the procedure.
COMPUTING A NEW VARIABLE (e.g., computing a new variable that is the sum of four other variables)
To compute a new variable from existing variables, you need to be in the SPSS data editor. Choose from the menus:
|
|
Transform |
|
|
Compute |
Type in the name of your new variable in the Target Variable box (e.g., Score) and go to the Numerical Expression box and type in or select the numerical expression that tells SPSS how the new variable will be created. For example, if the new variable (Score) is the sum of VarA, VarB, and VarC, the numerical expression would be: SUM(VarA, VarB, VarC). Then click on OK. SPSS will have created your new variable which will be located in the last column in your data file.
Handout prepared by
Dr. Kathleen Harring
Department of Psychology
Muhlenberg College
Revised 09/04/2006