SPSS command: SPSS MEANS

Group means and summary statistics by category

Q: How is MEANS different from One-Way ANOVA?

MEANS is a reporting procedure: many statistics, up to three crossed layers and an optional basic ANOVA. One-Way ANOVA focuses on the test, adding Levene, Welch, Brown-Forsythe, effect-size intervals and post hoc comparisons.

Q: What does the test for linearity show?

It splits the between-groups sum of squares into a straight-line trend (Linearity) and what is left (Deviation from Linearity). A significant Linearity row with a non-significant deviation means a straight line describes the group means well.

Q: How do I read eta squared?

It is the share of the outcome's variance explained by the first-layer groups: 0.06 is often called medium and 0.14 large. It is the same quantity as η² in One-Way ANOVA.

Q: Why is the ANOVA computed only for the first layer?

That is how SPSS MEANS works. For effects of several factors and their interaction, use GLM Univariate.

Q: Is my data sent anywhere?

No. MEANS runs on the data held in your browser tab; the file is not uploaded.

MEANS breaks a scale variable down by one or more categorical variables and reports the mean, N, standard deviation and other cell statistics for every group and subtotal.

Try with example data Run on your own file

Free · No account · Files stay on your device · Updated 18 October 2026

Worked example result

ANOVA Table: spend * tenure · Between Groups (Combined)

F(3, 44) = 12.432

p <.001 · R Squared = 0.416

Item	Sum of Squares	df	Mean Square
Between Groups (Combined)	2710.88	3	903.63
Linearity	2457.47	1	2457.47
Deviation from Linearity	253.41	2	126.71
Within Groups	3198.07	44	72.68
Total	5908.95	47

ANOVA Table: spend * tenure

See the full worked example

What it tells you

Crosstabs runs it in your browser with up to three crossed layers, 19 cell statistics including harmonic and geometric means, and an optional one-way ANOVA with eta and a test for linearity on the first layer.

The mean and spread of an outcome in every category, and in every combination of categories when layers are crossed.
Each group's share of the total N and of the total sum, useful for revenue or volume breakdowns.
With the ANOVA option, whether the first-layer groups differ and how much variance they explain (eta and eta squared).
With the test for linearity, whether the means follow a straight-line trend across numerically coded, ordered groups.

When to use it

Use it when

Reporting an outcome such as spend, score or time by segment, region or wave.
Breaking a mean down by two or three variables at once, such as region within gender.
A quick check of whether an ordered predictor such as age band relates linearly to the outcome.

Use something else when

Post hoc pairwise comparisons, Welch's test or effect-size intervals: use One-Way ANOVA.
Testing interactions between two factors: use GLM Univariate, since the MEANS ANOVA uses the first layer only.
A binary or categorical outcome: use a crosstab with row or column percentages.

How to run it in Crosstabs

01
Open your data in Statistics mode
Load the file in the workspace and switch to Statistics mode.
02
Choose Means
In the Compare means group, pick Means (MEANS).
03
Set the Dependent list and layers
Tick the scale variables in the Dependent list, then choose a grouping variable for Layer 1 of 3 and, if needed, Layers 2 and 3 to cross them. Leave Layer 1 as No grouping (whole sample) for overall statistics.
04
Pick the cell statistics
Under Cell statistics, Mean, N and Std. Deviation are on by default; add median, grouped median, standard error, sum, range, variance, skewness, kurtosis, harmonic or geometric mean, % of Total Sum or % of Total N.
05
Add the ANOVA and run
Tick ANOVA table and eta and, for numerically coded ordered groups, Test for linearity; then click Run.

Run Means in your browser

Worked example

Computed by the Crosstabs engine from the synthetic data below when this page was built — not typed in.

Try with example data

Data

48 synthetic customers, 12 in each of four tenure bands coded 1–4 with value labels; the dependent variable is monthly spend in dollars, drawn from normal distributions whose means rise with tenure but level off. Fixed seed.

Report: spend

tenure	Mean	N	Std. Deviation
Under 1 year	43.80	12	9.77
1–2 years	47.28	12	6.86
3–5 years	59.67	12	9.74
6+ years	61.00	12	7.31
Total	52.93	48	11.21

ANOVA Table: spend * tenure

Item	Sum of Squares	df	Mean Square	F	Sig.
Between Groups (Combined)	2710.88	3	903.63	12.432	<.001
Linearity	2457.47	1	2457.47	33.811	<.001
Deviation from Linearity	253.41	2	126.71	1.743	.187
Within Groups	3198.07	44	72.68
Total	5908.95	47

Measures of Association

Item	R	R Squared	Eta	Eta Squared
spend * tenure	0.645	0.416	0.677	0.459

How to read it

Mean spend differs across tenure bands, F(3, 44) = 12.43, p < .001, and η² = 0.459: tenure accounts for 45.9% of the variation in spend. The linear component is F = 33.81 (p < .001) while the deviation from linearity has p = .187; compare R² = 0.416 with η² to see how much of the group effect a straight line captures.

Output tables explained

Case Processing Summary: Included and excluded cases for each dependent variable; counts are unweighted.
Report: The chosen cell statistics for every group, with each level followed by its Total in SPSS layout.
ANOVA Table: Between Groups (Combined), Within Groups and Total sums of squares with F and significance; Linearity and Deviation from Linearity rows appear when the test for linearity applies.
Measures of Association: Eta and eta squared, plus R and R squared from the linear trend when linearity was tested.

Assumptions and notes

Cases with a missing value on the dependent or any layer variable are left out of that dependent's table.
The ANOVA F test assumes independent cases, roughly normal groups and similar variances; check them with Explore.
The test for linearity needs a numerically coded first layer with at least three groups; harmonic and geometric means need all values positive.
Under a sampling weight, the ANOVA, linearity test and standard errors are withheld; only weighted descriptive statistics are shown.

How the results are validated

Cell statistics are checked against SciPy 1.18 and NumPy 2.5, and the ANOVA, eta, linearity and deviation rows against statsmodels 0.15 OLS (anova_lm) on reference data, all to a relative tolerance of 1e-9. Frequency weights are tested to reproduce replicated cases, and user- and system-missing values to be excluded. Validation evidence

Questions

How is MEANS different from One-Way ANOVA?

MEANS is a reporting procedure: many statistics, up to three crossed layers and an optional basic ANOVA. One-Way ANOVA focuses on the test, adding Levene, Welch, Brown-Forsythe, effect-size intervals and post hoc comparisons.

What does the test for linearity show?

It splits the between-groups sum of squares into a straight-line trend (Linearity) and what is left (Deviation from Linearity). A significant Linearity row with a non-significant deviation means a straight line describes the group means well.

How do I read eta squared?

It is the share of the outcome's variance explained by the first-layer groups: 0.06 is often called medium and 0.14 large. It is the same quantity as η² in One-Way ANOVA.

Why is the ANOVA computed only for the first layer?

That is how SPSS MEANS works. For effects of several factors and their interaction, use GLM Univariate.

Is my data sent anywhere?

No. MEANS runs on the data held in your browser tab; the file is not uploaded.

Run Means on your own data

Open a .sav, CSV or Excel file. Everything runs in your browser; the file never leaves your device.

Open the workspace

SPSS is a trademark of IBM Corp. Crosstabs is not affiliated with or endorsed by IBM.