z-Test
Definition
A z-test examines statistical hypotheses using a standard normally distributed or suitably approximated test statistic.
A hypothesis test whose standardized test statistic follows or is sufficiently approximated by a standard normal distribution under the null hypothesis.
Simply explained
The test compares an observed deviation with what would be expected by chance under an assumption.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
An operation tests a defect rate in a large sample against a predefined reference value.
Limits and pitfalls
- Known variance or asymptotic assumptions must fit; significance does not replace effect size.
Origin & history
Z-tests originated from classical normal theory for means and proportions.
Use in mid-sized companies
- Possible example: An operation tests a defect rate in a large sample against a predefined reference value.
Frequently asked questions
What does z-test mean?
A hypothesis test whose standardized test statistic follows or is sufficiently approximated by a standard normal distribution under the null hypothesis. What should you watch out for? Known variance or asymptotic assumptions must be met; significance does not replace effect size.
Sources
- Tests of means · NIST SEMATECH e-Handbook, Tests of means
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: