Analyze alpha, tails, and critical thresholds instantly. Compare p-values, z-scores, and corrected limits with confidence. Turn hypothesis testing inputs into clear decisions and charts.
Enter a confidence level to compute alpha. Add a p-value for direct testing, or enter a z-score to estimate the p-value automatically.
Base significance level: α = 1 - confidence level
Bonferroni adjusted alpha: αadjusted = α / comparisons
Two-tailed critical z: zcritical = ± Z(1 - αadjusted / 2)
One-tailed critical z: zcritical = Z(1 - αadjusted) or Z(αadjusted)
Decision rule: Reject the null hypothesis when p-value ≤ alpha.
A significance level measures the probability of making a Type I error. Smaller alpha values create stricter decision thresholds.
When a z-score is supplied, the calculator estimates the p-value from the standard normal distribution based on the selected tail direction.
| Scenario | Confidence Level | Test Type | P-Value | Alpha | Decision |
|---|---|---|---|---|---|
| Marketing campaign lift test | 95% | Two-Tailed | 0.0320 | 0.0500 | Significant |
| Model accuracy improvement | 99% | Right-Tailed | 0.0180 | 0.0100 | Not significant |
| Feature drift alert check | 90% | Left-Tailed | 0.0710 | 0.1000 | Significant |
A significance level, often called alpha, is the chance of rejecting a true null hypothesis. It sets the cutoff for deciding whether a result looks statistically meaningful.
Alpha is the complement of the confidence level. For a 95% confidence level, alpha equals 0.05. For a 99% confidence level, alpha equals 0.01.
Tail direction changes the rejection region. Two-tailed tests split alpha across both tails, while one-tailed tests place all of it into one side of the distribution.
Use a p-value when you already computed it elsewhere. Use a z-score when you want this page to estimate the p-value and compare it against alpha automatically.
Bonferroni correction lowers alpha when many tests are run together. It helps reduce false positives by dividing the original alpha by the number of comparisons.
No. Statistical significance only shows whether a result crosses the chosen threshold. Practical importance still depends on effect size, context, cost, and real-world impact.
A multiple-testing correction makes the threshold stricter. A p-value that passes the original alpha can fail once the adjusted alpha becomes smaller.
Yes. It works well for A/B tests, model checks, drift analysis, feature screening, and many common validation tasks where alpha and p-values guide decisions.
Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.