Understanding Test Statistics
A test statistic turns sample evidence into one comparable number. It measures how far an observed result sits from a null claim. Larger distance usually means stronger evidence against that claim. This calculator supports common distributions used in classroom and research work. You can enter a statistic directly, or build it from summary values.
Why P Values Matter
A p value estimates how unusual the observed statistic would be when the null hypothesis is true. A small p value does not prove a claim. It only shows that the sample would be uncommon under the selected model. The tail choice changes the result. A right tailed test checks large positive statistics. A left tailed test checks small values. A two tailed test checks both directions.
Choosing the Right Test
Use a z test when the sampling distribution is normal, or a large sample supports normal approximation. Use a t test when a mean is tested with an estimated standard deviation. Degrees of freedom matter for t tests. Use chi square tests for variance, independence, or goodness of fit work. Use F tests for variance ratios and model comparisons.
Good Input Practices
Enter values from the same study design. Avoid mixing sample standard deviations, population deviations, and counts without checking the test type. Confirm the null value before calculating. Use positive sample sizes and valid degrees of freedom. When using proportions, enter counts for two proportion tests. Use decimal proportions for one proportion tests.
Reading the Result
The output includes the statistic, distribution, tail setting, p value, and decision at the chosen alpha level. If p is below alpha, reject the null hypothesis. If p is not below alpha, do not reject it. This wording is careful. It avoids saying the null is proven. Export options help save the result for reports, notes, worksheets, or later review.
Limitations
The calculator gives numerical support, not study design advice. Assumptions still matter. Independence, random sampling, normality, expected counts, and equal variance choices can affect conclusions. Always explain the practical meaning of the result, not only the final p value. Use it with course notes, field standards, and instructor guidance. Keep raw data available for verification and transparent review later checks.