ANOVAs and t-tests. One of the assumptions of ANOVAs is (like with t-tests) that the data are independent of each other. This is always going to be the case for our data – they are independent. As soon as comparisons are going to involve more than two sources of data it calls for an ANOVA – doing multiple t-tests is not encouraged as this increases the change of committing a type-1 error (coming up with a false positive). If you really want to apply multiple t-tests, then make sure to do a Bonferroni correction to adjust the probabilities.
Something else… may I suggest you install a R syntax highlighter here in your WordPress Blog? It’ll make your R code more readable. See here: http://www.viper007bond.com/wordpress-plugins/syntaxhighlighter/

Once things start to work nicely you can think about linking selected posts with R-Bloggers.

AJ

]]>