Find the critical value of z, t, chi-square, F or Pearson’s r for any significance level, one- or two-tailed. The graph shows the rejection region, so you can see where your test statistic needs to fall to reject the null hypothesis. Use it in place of z, t, χ², F and r tables.
Common critical values
| α | z (two-tailed) | z (one-tailed) | t, df = 10 (two-tailed) | χ², df = 1 |
|---|---|---|---|---|
| 0.10 | ±1.645 | 1.282 | ±1.812 | 2.706 |
| 0.05 | ±1.960 | 1.645 | ±2.228 | 3.841 |
| 0.01 | ±2.576 | 2.326 | ±3.169 | 6.635 |
Critical value vs p-value approach
Both reach the same decision. The critical-value approach compares your test statistic with a cut-off: reject H₀ if the statistic lands in the rejection region. The p-value approach compares the tail area beyond your statistic with α. Most software reports p-values, but critical values are still used to build confidence intervals and in exams. For p-values, use the p-value calculator.
Frequently asked questions
Why are there two critical values for a two-tailed χ² or F test?
These distributions are not symmetric, so the lower and upper cut-offs differ. Most χ² and F tests are right-tailed and need only the upper value. Two-tailed values are used, for example, in confidence intervals for a variance.
What is the critical value of r?
It is the smallest |r| that is significant for your sample size. With n = 25 and α = 0.05 (two-tailed) it is 0.396. Any correlation larger in magnitude is statistically significant.