SPSS command: SPSS CSDESCRIPTIVES

Survey means, totals and ratios with design-based errors

Complex Samples Descriptives estimates population means, sums, ratios and population sizes from a weighted, stratified or clustered sample, with standard errors that respect the design.

Try with example data Run on your own file

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

What it tells you

Crosstabs runs the SPSS CSDESCRIPTIVES procedure on your device, reporting t-based confidence intervals on the design degrees of freedom, coefficients of variation, design effects and unweighted counts, overall or by subpopulation.

The estimated population mean or total of each numeric variable, with a standard error from Taylor linearization over the declared strata and PSUs.
A ratio of two totals, such as spend per household member, with its own linearized standard error.
How much the design inflates or deflates the variance compared with simple random sampling (design effect and its square root).
Estimates for each category of a subpopulation variable, computed by domain estimation over the full design.
The estimated population size behind each estimate, next to the unweighted number of cases.

When to use it

Use it when

Reporting a survey mean or prevalence (a mean of a 0/1 variable) with an honest margin of error under clustering.
Estimating population totals (households, spend, hours) by weighting up a sample.
Comparing estimates across regions, age bands or other subgroups without discarding the cases outside each subgroup from the variance.

Use something else when

Your data are a simple random sample with no weight: Descriptives or Frequencies give the same answers more directly.
You want percentages of a categorical variable: use Complex Samples Crosstabs, which also tests associations.
You want to test whether subgroup means differ after adjusting for other variables: use Complex Samples GLM.

How to run it in Crosstabs

01
Open your survey file
Open the .sav, CSV or Excel file in the workspace and switch to Statistics mode. The file is read in your browser and is not uploaded.
02
Choose the procedure
In the Complex samples group, pick Complex Samples Descriptives (CSDESCRIPTIVES).
03
Check the saved sampling plan
The panel opens with a Sampling plan line showing the weight, strata, clusters and estimation. If no plan is saved for this dataset, open Sampling plan (CSPLAN) first. Plan variables are left out of the variable pickers, and the workspace weight is ignored: Complex Samples always use the plan's weight.
04
Choose measures and statistics
Tick numeric variables under Measures (numeric variables) and choose Mean, Sum and/or Population size. For a ratio, set Ratio numerator (optional) and Ratio denominator.
05
Add a subpopulation and options
Optionally pick a Subpopulation (optional) variable, set the Confidence level (90%, 95% or 99%) and under Missing values choose Exclude per analysis (SPSS default) or Exclude listwise across all variables. Then run.

Run Complex Samples Descriptives 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

A synthetic household survey of 120 respondents: three strata (Metro, Town, Rural), four primary sampling units (PSUs) per stratum and ten respondents per PSU, with a sampling weight that rises from about 35 in Metro to about 110 in Rural. Each PSU shares a random effect, so respondents in the same cluster resemble each other. The plan is weight + strata + clusters, with-replacement (WR) estimation; all values come from a fixed seed.

Means

Variable	sex	Estimate	Std. Error	95% CI Lower	95% CI Upper	Design Effect	Unweighted Count
spend	Total	172.27	6.56	157.43	187.11	2.820	120
spend	Female	182.55	5.77	169.50	195.59	0.876	59
spend	Male	164.21	8.10	145.90	182.52	2.846	61

Subpopulation estimates use domain estimation: cases outside the subpopulation contribute zero to the variance.

Ratios

Ratio	Estimate	Std. Error	95% CI Lower	95% CI Upper	Design Effect	Unweighted Count
spend / hhsize	58.528	3.151	51.399	65.656	1.742	120

Population Size

Variable	sex	Estimate	Std. Error	95% CI Lower	95% CI Upper	Unweighted Count
spend	Total	8308.2	72.9	8143.2	8473.2	120
spend	Female	3652.4	324.6	2918.1	4386.7	59
spend	Male	4655.8	307.0	3961.3	5350.3	61

Design: 3 strata, 12 primary sampling units; design df = 9. Intervals and tests use t or F with these df.

How to read it

The estimated population mean of spend is 172.27 (standard error 6.56, 95% CI 157.43 to 187.11 on 9 design df). Its design effect of 2.82 says the clustered, weighted sample is about as informative as 43 simple-random respondents rather than 120. Women's mean (182.55) and men's (164.21) are domain estimates over the full design, and the ratio estimate puts spend per household member at 58.53. The sample of 120 represents an estimated 8308 people.

Output tables explained

Means: Estimate, standard error, confidence interval, coefficient of variation, design effect, square-root design effect and unweighted count for each variable, by subpopulation category when one is chosen.
Sums: Estimated population totals with the same columns. Totals are weighted sums, so their standard errors also carry the uncertainty in the population size.
Ratios: The ratio of the numerator total to the denominator total, with a linearized standard error, interval and design effect, using the cases valid on both variables.
Population Size: The sum of the weights for the cases in each analysis or subpopulation, with its design-based standard error; no design effect is shown for population sizes.

Assumptions and notes

The saved plan describes how the sample was actually drawn: the right sampling weight, the stage-1 strata and the primary sampling units (PSUs). A wrong plan gives wrong standard errors, however good the model.
The sampling weight is used for estimation and the variance comes from the design; the weight is never treated as a frequency weight that replicates cases. Cases with a missing or non-positive weight, or a missing design variable, are excluded from the design.
Inference uses t and F distributions on the design degrees of freedom (PSUs minus strata). With few PSUs per stratum the df are small and intervals are wide; a stratum with a single PSU contributes no variance.
Cases missing an analysis variable, or outside a subpopulation, stay in the design with zero contribution, so every PSU still counts; cases missing the subpopulation variable itself are excluded.

How the results are validated

On a synthetic two-stage stratified survey (235 cases) under three designs — stratified clusters with replacement, two stages without replacement with population sizes, and stratified with sampling rates — totals, means, ratios and by-region domain means and totals match the R survey package (svytotal, svymean, svyratio, svyby) to a relative 1e-9. Design effects and confidence-interval half-widths are checked to the same tolerance against an independent NumPy/SciPy implementation of the SPSS formulas. Validation evidence

Questions

Why is my standard error larger than in ordinary Descriptives?

Respondents in the same cluster tend to resemble each other, so a clustered sample carries less information than its size suggests, and unequal weights add variance too. The design effect shows the factor: 2.5 means the variance is 2.5 times what a simple random sample of the same size would give.

Why not just filter the data to one subgroup?

Filtering can remove whole PSUs and strata and so understate the variance. The Subpopulation option uses domain estimation: cases outside the subgroup stay in the design with zero contribution, which is the standard approach for survey subgroups.

How do I estimate a proportion?

Code the characteristic as 0/1 and request its Mean: the mean of a 0/1 variable is the population proportion, with a design-based standard error. For a full categorical distribution with logit intervals, use Complex Samples Crosstabs.

Can I compute several ratios at once?

The panel takes one numerator and one denominator per run, alongside any number of means and sums. Run it again with another pair for a second ratio.

Is my survey file uploaded?

No. The workspace reads the file and runs CSDESCRIPTIVES in your browser, and the sampling plan is stored in this browser only. Nothing is sent to Crosstabs servers, and there is no account to create.

Run Complex Samples Descriptives 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.