P-value is the probability of obtaining extreme results under null hypothesis. Null hypothesis states no difference between observed and expected values. Alternative hypothesis proposes difference between observed and expected values
Statistical test for paired nominal data in 2x2 contingency tables. Named after Quinn McNemar, introduced in 1947. Determines equality of row and column marginal frequencies
Fisher's exact test is a statistical significance test for contingency tables. Test was developed by Ronald Fisher based on Muriel Bristol's tea tasting experiment. Valid for all sample sizes, unlike many other exact tests
Chi-square tests are nonparametric statistical tests for categorical data. Test statistic (Χ²) compares observed and expected frequencies. Categorical variables represent groupings like species or nationalities
P-value measures likelihood of observing data under null hypothesis. Smaller p-value indicates stronger evidence for alternative hypothesis. Census Bureau requires p-values above 0.10 to be accompanied by zero comparison
P>.05 indicates that null hypothesis should not be rejected. Type-i errors occur when rejecting null due to sampling error. P-values measure probability of random samples differing from null