Ask the Professor Forum

Hi Javier,

I created a video for this problem for you.  You wanted me to work out all of the sum of squares and that takes a while, so the video is long. I left out some of the details on comparing the test stats to the critical values because I assume you know how to do that from the other sections of chapter 10.

Just remember that we first test if there is a treatment effect (Ho: No difference among the ab treatment means). If we reject that null, then we test for an interaction effect (Ho: Factors A and B do not interact). If we reject that (i.e. we show there is an interaction effect), we do not need to test if A and B are significant separately. If you cannot show there is an interaction effect, then you would test A and B separately.

Hope the video helps,

Professor