Unlike many experimental situations, with computer modelling the problem
tends to be an embarrassment of riches. There is no problem with amount of
data. You can gather scads and scads of data. So much so that it can be
hard to find statistical tests that deal with what you’ve got. Right now,
I have data that can be set up and analysed as a randomized block (or
repeated measures) experiment.
Here is an excellent description of randomized block ANOVA and how it
works. The problem is, I not only have randomized blocks, I have multiple
replicates within each block. Psychologists should be so lucky, right? But
no one seems to cover this case, because usually people are lucky to get
enough subjects to do a randomized block experiment with a single value in
each treatment of each block. So, what to do now? I could take the average
of the values and test those, or I could randomly pick one of my
replicates to compare. Both seem plausible. But what is the ‘officially
right’ thing to do? That’s my dilemma.
posted at 11:34 on Mon, 23 Jun, 2008 |
path: /school
All material copyrighted by Jen Schellinck. All rights reserved.