![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Statistical Testing
Links:
Univariate & Bivariate Methods
Univariate methods include tests that compare similar variables for different groups. Two groups can be compared, e.g., biological measures at reference or impacted sites. When more than two groups are compared, ANOVA can be used to make comparisons. Bivariate methods typically include correlation and regression, 2 similar approaches for testing whether 2 variables track each other.
Multivariate Methods
Multivariate methods are easiest to understand in terms of the numbers of independent and dependent variables included in the model. Some methods yield a p-value for significance, e.g., MANOVA, while others are exploratory, e.g., PCA.
Distribution-Free Methods
Distribution-free methods include nonparametric tests based on ranks as well as computer-intensive techniques that build new distributions from models or data.
Comparisons of Statistical Tests
Tables and comparisons found here include nonparametric equivalent tests, pair-wise comparison of all the statistical tests presented in this web site, and a comparison of the underlying models for multivariate tests.