Lesson 06.04 · Workbook pages 616–622
Run the ANCOVA model
You have finished the assumption checks. Now you will place the dependent variable, grouping variable, and covariate into SPSS, request only the needed options, run the model, and verify the group structure before interpreting results.
The goal of this lesson
Generate the ANCOVA output correctly, then pause at the Between-Subjects Factors table. That first table checks your setup. It is not a significance test.
Step 1 and Step 2
Open the model and place the variables
ANCOVA uses the same SPSS procedure as ANOVA, but it adds a continuous covariate. The workbook scenario compares PostTest scores across Program groups while controlling for Pretest.
Keep the three roles separate. The outcome goes in Dependent Variable, the grouping variable goes in Fixed Factor(s), and the control variable goes in Covariate(s).
Dependent VariablePostTest
Fixed Factor(s)Program
Covariate(s)Pretest
| SPSS box | Variable | What the role means |
|---|---|---|
| Dependent Variable | PostTest | The continuous outcome being compared. |
| Fixed Factor(s) | Program | The categorical grouping variable. |
| Covariate(s) | Pretest | The continuous variable being controlled. |
Step 3 and Step 4
Request the needed options, then run
The workbook keeps the setup deliberately simple. Click Options, choose the requested output, continue, and click OK.
1
Open OptionsClick Options… in the Univariate dialog.
2
Select descriptivesChoose Descriptive statistics.
3
Effect size, when consistentSelect Estimates of effect size only when effect size has been included consistently in the course.
4
Run the modelClick Continue, then OK.
✓Descriptive statistics
Selected for the workbook run.
✓Estimates of effect size
Selected only when your earlier ANOVA lessons used effect size consistently.
Novice guardrail
Do not select extra options just because they are available. The workbook asks you to keep the run simple so each output can be interpreted one piece at a time.
Output 1
Read the Between-Subjects Factors table
SPSS produces this table automatically. It confirms that the program groups are coded correctly and that the sample sizes match the intended design. It does not test a hypothesis.
Between-Subjects Factors
The workbook scenario has three balanced program groups.
| Numeric code | Program label | N |
|---|---|---|
| 1 | After-School Tutoring | 20 |
| 2 | Saturday Academy | 20 |
| 3 | Summer Bridge | 20 |
The model is comparing the three intended program categories.
The sample is balanced across the program groups.
The table represents a between-subjects design with distinct group membership.
Codes 1, 2, and 3 correspond to the intended program names.
What you may conclude now
The ANCOVA is set up to compare the intended groups. You have not yet learned whether the groups differ. Output 1 only verifies the structure.
Stop and check
Fix the setup before interpreting anything
ANCOVA can adjust outcome scores. It cannot repair an incorrect grouping structure.
!
The group labels are wrong, swapped, or inconsistent with the study.
!
A program group is missing or unexpectedly small.
!
Participants appear to belong to more than one group.
!
The sample sizes do not match what you expected from the dataset.