Hi everyone,
I just had a quick question about how to interpret my Brain Scores Plot (with CI). I have a two-group design with three conditions. I've run a non-rotated task analysis, where I specified a -0.5 -0.5 1 contrast for my three conditions for both groups to look at the main effect of task (where Condition 3 is the one I'm interested in, and Conditions 1 and 2 are my control tasks).
This is what my Brain Scores Plot looks like for this analysis (where the second figure is just a re-drawn graph with the bars are grouped by condition, rather than group): https://www.dropbox.com/s/0hon62sqlwdneiv/fig.docx?dl=0
From this (which was significant), it seems that the CIs overlap between Conditions 2 and 3 (which indicates that there may be no significant differences across these two conditions). However, my understanding is that when the two control tasks are summed and weighted against the third experimental condition, they average out, and it's the difference between this average and Condition 3 that is significant - is this correct? Am I able to use this to state that there is a main effect of task? Or would I have to run another analysis that includes just Conditions 2 and 3? Would I do this by leaving out Condition 1 entirely, or weight the conditions like this: 0 -1 1?
Thank you so much!
Haeme
Hi Haeme - sorry for the delay - vacation time. You are correct that in order to address whether 2 & 3 differ you will need to enter it explicitly as a contrast (0 1 -1). You can say there is a task main effect, as per your contrast, but you know that it is being driven by the diffs between 1 & 3
Baycrest is an academic health sciences centre fully affiliated with the University of Toronto
Privacy Statement - Disclaimer - © 1989-2024 BAYCREST HEALTH SCIENCE. ALL RIGHTS RESERVED