Critical Values and P-values for Statistical Tests
There are two approaches to conducting significance tests. Some analysts like to compare the test statistic with the critical value for a given significance level; others prefer to calculate the P-value corresponding to the test statistic. Excel can be used for either method. Assuming significance level $$\alpha$$, (typically $$\alpha$$ = 5% or 0.05):
| Two-tailed z-test | ||
|---|---|---|
| Upper tail critical value | =NORMSINV(1-alpha/2) | |
| P-value for given z | =2*(1-NORMSDIST(ABS(z))) | |
| Two-tailed t-test with v degrees of freedom | ||
| Upper tail critical value | =TINV(alpha, v) | |
| P-value for given t | =TDIST(ABS(t), v, 2) | |
| One-tailed $$\chi^2$$-test with v degrees of freedom | ||
|---|---|---|
| Upper tail critical value | =CHIINV(alpha, v) | |
| P-value for given chisquared | =CHIDIST(chisquared, v) | |
| One-tailed F-test with v1 degrees of freedom in the numerator and v2 in the denominator | ||
| Upper tail critical value | =FINV(alpha, v1, v2) | |
| P-value for given F | =FDIST(F, v1, v2) | |
