Post-hocs in a 2-way ANOVA...

#1
I have a setup in which cell viability is dependent on materials and time. A two-way ANOVA gives all within- and between-subject comparison P values. But I need to compare cell viability on each material with its competitors at each interval. Also I need to compare the cell viability on each material through time (comparison between different intervals for each material).

I have read several tutorials about this issue. Using post-hoc tests in such a design will compare "only" time intervals with each other, and each material with the others. I mean when it is comparing the time intervals, it compares "all" materials at each interval with "all" materials at other intervals. When comparing materials, it now combines the time intervals before between-material comparisons!

Is there any other program that can give me the specific tests I need? Currently I have to do two other one-sample ANOVAs to calculate the pairwise comparisons using their Tukeys. However I know it is incorrect, but no other software I know which can do it.
 

trinker

ggplot2orBust
#2
I'm out the door right now but R can do this for you. Hopefully others cime in here. You may want to ask a moderator to move this thread to the R forum (by the way R is free). One caveat, R has a little bit of a learning curve but the pay off is huge.
 

Dason

Ambassador to the humans
#3
There really isn't a definitive GUI for R. And really most people that eventually learn R think that's alright. A programming language has a steeper learning curve but is a lot easier to use once you learn it and it is quite a bit more powerful than trying to do everything menu based. But with that said there are a few programs that try to offer up a GUI.

Rcmdr and Rkward are two that I know of that offer some menu driven capabilities. RStudio is nice but it's more of an IDE and doesn't really do any analysis through menus - it just offers a nicer environment to work with R in.

With that said I don't know really know much about SPSS but it's probably possible to do what you want to do in SPSS?
 
#4
Hi Victorxstc,

I am running a 2way repeated measures anova. In additional to the fact that SPSS does not do post hoc on interactions, I have also read that SPSS posthoc on more than 1 way should not be used. Is contrast , a more appropriate thing to do?