estimates parameters of a linear model that contains effects of treatments (and perhaps their interactions etc.) and residual effects representing random variation. The analysis-of-variance-table partitions the sum of squares of the data values into the components that can be explained by treatment terms and those arising from random variation. The V.R. and F.prob columns allow you to assess whether the treatment effects are larger enough not to have been caused by random variability.
Updated on December 4, 2017