Which table do you need?
| If you are… | Use |
|---|---|
| Finding a percentile or probability for a normal variable, or a z critical value | Z-table |
| Running a t-test or building a confidence interval for a mean from a sample | T-table |
| Testing independence in a contingency table, or goodness of fit | Chi-square table |
| Checking an ANOVA or regression F statistic, or comparing two variances | F-table |
| Comparing pairs of group means after a significant ANOVA | Tukey HSD (q) table |
| Comparing two independent groups with ranks (small samples) | Mann–Whitney U table |
| Comparing paired measurements with ranks (small samples) | Wilcoxon signed-rank table |
| Finding the probability of k or fewer successes in n trials | Cumulative binomial table |
| Testing whether a correlation coefficient differs from zero | Pearson r table |
Critical values versus p-values
A table gives a critical value: the cut-off your test statistic must pass to be significant at a chosen level such as 0.05. Software gives an exact p-value instead. Both lead to the same decision, but the p-value shows how strong the evidence is, so most journals ask for it. Tables are still the quickest way to check a result by hand, and they are expected in many exams. Each table page explains how to read it, with a worked example.
How these tables were made
Every value is computed with SciPy's distribution functions rather than copied from older printed tables, so there are no transcription errors. Each page names the SciPy function used. If you need a value that is not in a table, such as an unusual degrees of freedom, the matching calculator gives the exact figure.