SPSS command: SPSS GRAPH /ERRORBAR

Error bar charts in your browser

An error bar chart shows each group's mean with a bar for its uncertainty or spread.

Run Error Bars in your browser All SPSS procedures

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

What it tells you

Crosstabs draws SPSS-style error bars for a confidence interval of the mean, a multiple of the standard error, or a multiple of the standard deviation, and lists the numbers behind every point.

Each group's mean and how precisely it is estimated (confidence interval or standard error bars).
How spread out the individual values are within each group (standard deviation bars).
The exact N, mean, SD, SE and bar ends per group, for reporting.

When to use it

Use it when

Presenting group means from a survey or experiment alongside an ANOVA or t test.
A quick visual check of which group means are estimated precisely and which rest on few cases.

Use something else when

As a significance test: overlapping 95% bars do not prove means are equal. Run One-Way ANOVA or a t test.
Survey data with a sampling weight when you need intervals: CI and SE bars are withheld; use Complex Samples Descriptives.
Heavily skewed data, where a mean ± SD bar can run below the minimum; show medians in Explore instead.

How to run it in Crosstabs

01
Open your data
Open a .sav, CSV or Excel file in the workspace and switch to Statistics mode.
02
Choose Error Bars
In the Utilities and charts group pick Error Bars, choose the Variable (scale) and, optionally, a Category axis (optional) variable for the groups.
03
Pick what the bars represent
Bars represent: Confidence interval for the mean (Level 90%, 95% or 99%), Standard error of the mean or Standard deviation, each with a Multiplier (default 2).
04
Run
Declare an active weight as frequency or sampling, then press Run to draw the chart and the Error Bar Statistics table.

Run Error Bars in your browser

Worked example

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

Data

Satisfaction scores (0–10) from 48 synthetic shoppers in three store formats with unequal group sizes (24, 9 and 15), drawn with a fixed seed.

Error Bar Statistics (95% CI of the mean)

Item	N	Mean	Std. Deviation	Std. Error	Bar lower	Bar upper
Mall	24	7.263	1.101	0.225	6.798	7.727
Outlet	9	6.711	1.265	0.422	5.738	7.684
Street	15	6.320	1.360	0.351	5.567	7.073

These rows are the points and whiskers the chart draws. Intervals use Student's t with N − 1 df.

Error Bar Statistics (Mean ± 2 SE)

Item	Mean	Bar lower	Bar upper
Mall	7.263	6.813	7.712
Outlet	6.711	5.868	7.555
Street	6.320	5.618	7.022

How to read it

Outlet has the widest 95% bar (5.74 to 7.68) because the interval width grows with the standard deviation and shrinks with √N, and it has only 9 shoppers; Mall has the narrowest (6.80 to 7.73). Overlapping confidence bars do not by themselves prove the means are equal: use One-Way ANOVA for the test.

Output tables explained

Error Bars: The chart: one point per group at its mean, with whiskers from the bar lower to the bar upper value.
Error Bar Statistics: N, mean, standard deviation, standard error, bar lower and bar upper for each group, exactly as plotted.

Assumptions and notes

Confidence intervals use Student's t with W − 1 degrees of freedom, W the sum of weights, and so assume roughly normal means.
Cases with a missing value on the variable are excluded per variable, and cases with a zero or missing weight are left out.
With a sampling weight, means and SDs are weighted but SE and CI bars are blocked, because they would assume independent cases.

How the results are validated

Error-bar statistics are checked against an independent Python reference: weighted mean, SD and SE agree to 1e-12 and confidence limits, computed with scipy.stats.t quantiles, agree to 1e-10 at each tested level. The engine's median and its Hettmansperger–Sheather interval are also checked against scipy.stats.binom to 1e-10, though the panel currently charts means only. Validation evidence

Questions

Should I show confidence intervals or standard errors?

Confidence intervals are easier to read correctly and adjust for group size through the t quantile. Mean ± 1 SE bars are narrower than 95% CIs and are often misread; state clearly what the bars show.

If two 95% bars overlap, are the means not different?

Not necessarily. Two means can differ significantly even when their 95% intervals overlap a little. Use a t test or ANOVA for the comparison.

Why are my CI bars disabled?

You declared the weight as a sampling weight. Choose standard deviation bars, or use Complex Samples Descriptives for design-based intervals.

Can I export the chart's numbers?

Yes. The Error Bar Statistics table lists every plotted value and can be copied or downloaded like other output tables.

Run Error Bars 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.