# Hello, I need some help on starting this assignment, I don't know how to do the anova test for question one and whether it should be one or two way.

Hello, I need some help on starting this assignment, I don't know how to do the anova test for question one and whether it should be one or two way.

There's a good explainer here and a short textbook with everything you'd want to know starting out.

In general, an `ANOVA` takes the ratio of the between sample variance to the within sample variance as a basis for comparison to the Fisher probability distribution to answer the question are the means of these two variables different at some predetermined confidence interval? The null hypothesis is that they are not; if the test statistic exceeds the selected p-value, the null is rejected in favor of the alternative—that they are different. For multiple pairs of variables, with as few as 3 tests being considered, there is a 0.1426 chance of observing at least one significant result by chance, even if all of the tests are actually not significant. Approaches 0.99 with 14 groups. Post-hoc tests adjust to give a “global” significance level.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.