Binomial test
Compares the proportion of one of two outcomes (yes/no, pass/fail) with a test proportion, using the exact binomial distribution.
Menu: Analyze → Nonparametric Tests → Legacy Dialogs → Binomial…
When should I use it?
- Is the pass rate different from 50%?
- Is the share of defective items above the 5% target?
What are the assumptions?
- A dichotomous variable, or a scale variable split at a cut point.
- Independent cases.
How do I run it in StatLab?
- Move the variable to the test list.
- Set the test proportion (default .50).
- For a scale variable choose 'Cut point' — values at or below it form group 1.
How do I read the output?
| Output | Meaning |
|---|---|
| Group 1 / Group 2 | Group 1 is the first value met in the data (or ≤ cut point); the test proportion refers to Group 1. |
| Exact Sig. (2-tailed) | Shown when the test proportion is .50. |
| Exact Sig. (1-tailed) | Shown for any other test proportion, in the direction of the observed result. |
Worked example
On the practice survey dataset (File → Sample dataset), this command — which the dialog's Paste button also produces — runs the procedure:
PROC npBinomial /vars=(smoker) /p=0.5 /define=data.
Output tables: Binomial Test.
How would I report it in APA style?
An exact binomial test was used. Smoker: 104 of 150 (69.3%) in group 1 vs. test proportion .50, exact p < .001 (2-tailed).
Computed by StatLab's validated procedure code when this page was built. The example only illustrates the steps; choose variables that fit your own research question.
What are common mistakes?
- Forgetting that the test proportion applies to Group 1, which depends on the data order.