SPSS tutorials website header logo SPSS TUTORIALS BASICS ANOVA REGRESSION FACTOR CORRELATION

SPSS ANOVA without Raw Data

In SPSS, you can fairly easily run an ANOVA or t-test without having any raw data. All you need for doing so are

of the dependent variable(s) for the groups you want to compare. This tutorial walks you through analyzing the journal table shown below.

SPSS ANOVA Without Data Means Table

1. Set Up Matrix Data File

First off, we create an SPSS data file containing 3 rows for each group. You may use descriptives-for-anova-01.sav -partly shown below- as a starting point.

SPSS Matrix Data File Example

There's 3 adjustments you'll typically want to apply to this example data file:

SPSS Matrix Data File Variable View

I recommend you don't make any other changes to this data file or otherwise the final analysis is likely to crash. For instance,

2. SPSS Oneway Dialogs

First off, make sure the example data file is the only open data file in SPSS. Next, navigate to Analyze SPSS Menu Arrow Compare Means SPSS Menu Arrow One-Way ANOVA and fill out the dialogs as if you're analyzing a “normal” data file.

SPSS ANOVA On Descriptive Statistics Dialogs

You may select all options in this dialog. However, Levene's test -denoted as Homogeneity of variances test- will not run as it requires raw data.

This is no problem for the data at hand due to their equal sample sizes. For other data, it may be wise to carefully inspect the Welch test as discussed in SPSS ANOVA - Levene’s Test “Significant”.

Anyway, completing these steps results in the syntax below. But don't run it just yet.

*WRONG SYNTAX AS PASTED FROM MENU.

ONEWAY Value BY Group
/ES=OVERALL
/STATISTICS DESCRIPTIVES WELCH
/PLOT MEANS
/MISSING ANALYSIS
/CRITERIA=CILEVEL(0.95)
/POSTHOC=TUKEY ALPHA(0.05).

3. Adjusting the Syntax

Note that we created syntax just like we'd do when analyzing raw data. You could run it, but SPSS would misinterpret the data as 4 groups of 3 observations each. For SPSS to interpret our matrix data correclty, add /MATRIX IN(*). as shown below.

*CORRECTED SYNTAX FOR SPSS 27 OR HIGHER.

ONEWAY Value BY Group
/ES=OVERALL
/STATISTICS DESCRIPTIVES WELCH
/PLOT MEANS
/CRITERIA=CILEVEL(0.95)
/POSTHOC=TUKEY ALPHA(0.05)
/matrix in(*).

*CORRECTED SYNTAX FOR SPSS 26 OR LOWER.

ONEWAY Value BY Group
/STATISTICS DESCRIPTIVES WELCH
/PLOT MEANS
/CRITERIA=CILEVEL(0.95)
/POSTHOC=TUKEY ALPHA(0.05)
/matrix in(*).

4. Interpreting the Output

First off, note that SPSS has understood that our data file has 4 groups of 25 observations each as shown below.

SPSS ANOVA Without Raw Data Output

The remainder of the output is nicely detailed and includes

The main output you cannot obtain from these data are

For the sample sizes at hand, however, none of these are very useful anyway. A more thorough interpretation of the output for this analysis is presented in SPSS - One Way ANOVA with Post Hoc Tests Example.

Right, so I hope you found this tutorial helpful. We always appreciate if you throw us a quick comment below. Other that that:

Thanks for reading!

Tell us what you think!

*Required field. Your comment will show up after approval from a moderator.

THIS TUTORIAL HAS 9 COMMENTS: