SPSS command: SPSS RELIABILITY

Cronbach's alpha and scale reliability

Reliability analysis checks whether the items of a multi-item scale measure the same thing consistently enough to be summed or averaged.

Try with example data Run on your own file

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

Worked example result

Cronbach's Alpha = 0.824

Item	Scale Mean if Item Deleted	Scale Variance if Item Deleted	Corrected Item-Total Correlation
q1	11.53	10.931	0.724
q2	11.65	11.011	0.742
q3	11.75	11.716	0.660
q4	11.72	10.851	0.758
q5	11.88	14.715	0.244

Item-Total Statistics

See the full worked example

What it tells you

Crosstabs runs SPSS RELIABILITY in your browser with alpha, split-half, Guttman, parallel, strictly parallel and McDonald's omega models, item-total statistics, ANOVA-based tests, Hotelling's T² and intraclass correlation coefficients.

How internally consistent the scale is: Cronbach's alpha, its standardized version, or omega under a single-factor model.
Which items do not fit: a low corrected item-total correlation, or an alpha that would rise if the item were deleted.
Possibly reverse-keyed items, whose corrected item-total correlation is negative.
Whether item means differ (F, Friedman or Cochran test) and, with ICCs, how well raters or measurements agree.

When to use it

Use it when

Before summing Likert items into a scale score, to report its reliability.
Shortening a questionnaire by spotting items that add little.
Measuring agreement among raters scoring the same cases, with an intraclass correlation.

Use something else when

To show that items measure a single construct: alpha assumes it; run Factor analysis to check dimensionality.
Agreement between two raters on categories: use Cohen's kappa or weighted kappa.
Items with reverse-worded statements that have not been recoded: recode them first or alpha will be understated.

How to run it in Crosstabs

01
Open the item data
Load the file with one column per item and switch to Statistics mode.
02
Choose Reliability analysis
In the Scale group, pick Reliability analysis (RELIABILITY) and tick two or more Items.
03
Pick the Model
Choose Alpha (Cronbach), Split-half (set the items in part 1), Guttman, Parallel, Strict parallel or Omega (McDonald).
04
Tick Statistics and the ANOVA table
Item descriptives, Scale descriptives, Scale if item deleted and Summaries are on by default; add Inter-item correlations or covariances, Hotelling's T-squared or Tukey's test of additivity, and choose an ANOVA table: F test, Friedman chi-square or Cochran chi-square.
05
Add an intraclass correlation if needed
Tick Intraclass correlation coefficient and choose Two-way mixed, Two-way random or One-way random, Consistency or Absolute agreement, the confidence interval and test value. Results update as you change options; declare any weight first.

Run Reliability Analysis 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

60 synthetic respondents answering five 1–5 items of a workload scale. Items q1–q4 load strongly on one underlying trait; q5 barely does. Generated with a fixed seed and rounded to whole scale points.

Reliability Statistics

Cronbach's Alpha	Cronbach's Alpha Based on Standardized Items	N of Items
0.824	0.820	5

Item-Total Statistics

Item	Scale Mean if Item Deleted	Scale Variance if Item Deleted	Corrected Item-Total Correlation	Squared Multiple Correlation	Cronbach's Alpha if Item Deleted
q1	11.53	10.931	0.724	0.588	0.757
q2	11.65	11.011	0.742	0.655	0.752
q3	11.75	11.716	0.660	0.509	0.777
q4	11.72	10.851	0.758	0.640	0.746
q5	11.88	14.715	0.244	0.093	0.883

Summary Item Statistics

Item	Mean	Minimum	Maximum	Range	N of Items
Item Means	2.927	2.750	3.100	0.350	5
Inter-Item Correlations	0.476	0.144	0.730	0.587	5

How to read it

Cronbach's alpha for the five items is 0.824 (standardized 0.820), with a mean inter-item correlation of 0.476. Item-Total Statistics shows the weak item: q5 has a corrected item-total correlation of 0.244, and dropping it would raise alpha to 0.883. Check the item's wording before deleting it; alpha alone does not prove the remaining items measure one construct.

Output tables explained

Reliability Statistics: For the alpha model, Cronbach's Alpha, alpha based on standardized items and N of Items; other models show split-half coefficients, Guttman's lambdas 1–6, parallel-model variances or McDonald's omega.
Item Statistics: Mean, standard deviation and N for each item.
Summary Item Statistics: Mean, minimum, maximum, range, max/min ratio and variance of the item means, item variances and inter-item covariances and correlations.
Item-Total Statistics: Scale mean and variance if the item is deleted, corrected item-total correlation, squared multiple correlation and alpha (or omega) if the item is deleted.
ANOVA: Between People, Between Items and Residual sums of squares with F, Friedman's chi-square or Cochran's Q, plus Tukey's nonadditivity test when requested.
Intraclass Correlation Coefficient: Single and average measures ICC with confidence intervals and the F test against the chosen test value.

Assumptions and notes

Cases are complete on every item: listwise deletion is applied and items with zero variance are dropped.
Alpha assumes the items measure one dimension with roughly equal loadings (tau-equivalence); when loadings differ, omega is the better estimate.
Under a sampling weight, reliability coefficients are reported but significance tests are withheld.

How the results are validated

Checked against pingouin 0.7, NumPy 2.5 and R 4.6.1 psych 2.6.9: alpha, standardized alpha, corrected item-total correlations and alpha-if-deleted agree within 1e-12, split-half coefficients and Guttman's lambdas within 1e-12, the ANOVA and Tukey nonadditivity F within 1e-9, Friedman (SciPy 1.18) and Cochran's Q (statsmodels 0.15, pingouin) within 1e-9, and ICCs within 1e-12 of R psych::ICC and pingouin. Frequency weights reproduce replicated cases. Validation evidence

Questions

What is a good Cronbach's alpha?

Common rules of thumb call 0.70 acceptable, 0.80 good and 0.90 excellent. Very high values (above about 0.95) can mean redundant items rather than a better scale.

Should I delete the item that raises alpha?

Only if it is weak for substantive reasons too. Check its wording and its corrected item-total correlation; a gain of a few hundredths is rarely worth losing content coverage.

Alpha or omega?

Omega allows items to load unequally on the common factor, so it is usually the more accurate estimate. The Omega (McDonald) model reports it with omega-if-item-deleted; it is withheld when a single-factor model does not fit, often because an item is reverse-keyed.

Which ICC model should I choose?

Use two-way random with absolute agreement when raters are a sample from a larger pool and their exact scores matter; two-way mixed when these specific raters are the only ones of interest; one-way random when each case has different raters.

Is my questionnaire data uploaded?

No. Reliability is computed in your browser and the data stays on your device.

Run Reliability Analysis 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.