Reliability (Cronbach's alpha)
Reliability analysis checks whether a set of questionnaire items measures the same thing consistently. Cronbach's alpha above about .70 is usually considered acceptable.
Menu: Analyze → Scale → Reliability Analysis…
When should I use Reliability?
- A multi-item scale (e.g. five satisfaction items) before you add the items into a total score.
What are the assumptions?
- Items scored in the same direction — reverse-score negatively worded items first.
- Cases with any missing item are excluded (listwise).
How do I run it in StatLab?
- Choose two or more items.
How do I read the output?
| Output | Meaning |
|---|---|
| Cronbach's Alpha | .9+ excellent, .8 good, .7 acceptable, .6 questionable, below .5 unacceptable (George & Mallery rule of thumb). |
| Based on Standardized Items | Alpha computed from correlations; differs when item variances differ. |
| Corrected Item-Total Correlation | Correlation of each item with the sum of the others; below about .30 means the item fits poorly. |
| Alpha if Item Deleted | If alpha would rise clearly without an item, consider removing or revising it. |
Worked example
Items q1–q5 in the bundled sample dataset (File → Sample dataset: 150 survey respondents with gender, programme group, age, pre/post test scores, five satisfaction items, income, smoking and education): α = .87 (n = 150).
How would I report it in APA style?
The five-item satisfaction scale showed good internal consistency, Cronbach's α = .87.
Numbers above were computed by StatLab’s validated engine from the sample dataset when this page was built.
What are common mistakes?
- Forgetting to reverse-score negatively worded items (gives a low or even negative alpha).
- Treating a high alpha as proof that the scale measures only one construct — that needs factor analysis.