In an ANOVA analysis, how can I calculate SS s/a or MS s/a without having individual observations, and only having a mean score (and no standard deviation) for each group?
Background: I mentioned last week in a thread about something else that I was looking at a problem in an experimental design book in which certain calculations were made that I could not understand how they were done because individual observations was missing. This is not homework, by the way. I’m trying to understand how the book is able to arrive at answers without what appears to be vital information.