ANOVA relies on an F-distribution as the basis of all probability distribution. An F score is obtained by dividing the between-group variance by the in-group variance. If the overall F test was significant, you can conduct a post hoc test (or multiple comparison tests) to measure the differences between groups. The most commonly used post hoc tests are Scheffé's method, the Tukey-Kramer method, and the Bonferroni correction.

In order to interpret the output of ANOVA, you need to have a basic understanding of certain terms, including the degrees of freedom, the sum of square total, the sum of square groups, the sum of square errors, the mean square errors, and the F statistic. If you require more information about these terms, you may refer to Using Multivariate Statistics (Fidell, L. S., and Tabachnick, B. G. (2006) Boston: Allyn & Bacon.), or refer to the Wikipedia entry of Analysis of Variance (http://en.wikipedia.org/wiki/Analysis_of_variance#cite_ref-31).