🚀 go-pugleaf

Could someone well versed in the assumptions underlying ANOVA please
read this and give me their opinion on whether what I did is OK and
interpretable?

Samples located in a three dimensional array of colors were named by 
subjects one at a time.  For each color name, a centroid was determined 
by averaging the coordinates of the samples named by that color.  
This was done in three different conditions, so that there were 11 
centroids for each condition.  

For each pair of conditions (A:1 vs 2; B:1 vs 3; C:2 vs 3), I calculated the 
distance (in absolute value) that each of the ll  centroids changed for each 
subject.  These difference distances were then used in an ANOVA where one 
of the factors was condition-pair (A, B, or C).  The hypothesis is that a large change
in location of the centroids in one of the conditions would lead to a 
significant effect of condition-pair.  More specifically if conditions 2 and
3 were similar and 1 was different from them both, you would also get
a significant effect of condition pair between A and C, and between B and C,
but not A and B since conditions 2 and 3 are similar in effect.

This is what I found but my question is that since the values are not
independent is this analysis no good.  Can I present it with the caveat
that they are not independent?  Does the large df I have help?
Presented graphically, the result looks good, but journals like to
have statistics to go with the pictures.


Please e-mail me since I don't think others will
be too interested.

Thanks,
Ethan Montag

em@cvs.rochester.edu
 ____              ___                             ___     ___    ___   ___
|\___\_   _____   |\__\                           |\___\_  \__\  /\__\_|\__\
| |  |_\ |\_____\_| |  | Ethan D. Montag          | |  |_\  \  \/ /  /\_\|  |
| |     || |     |_\|  | Center for Visual Science| |     |  \  \/  /| |    |
| |     || | ___       | University of Rochester  | |     |   \    / | |    |
| |     | \|_|| |      | 274 Meliora Hall         | |     |    \  /  | |    |
| |     | __  | |      | Rochester, NY  14627     | |     ||\   \//\ | |    |
| |     ||\_\_| |      |                          | |     || |\  / /|| |    |
| |     || | |_\|      | e-mail:                  | |     || | \/ / || |    |
| |     || |     ___   |  em@cvs.rochester.edu    | |     || |  \/  || |    |
| |    _| \|_____|| |  | Phone: (716)275-8443     | |    _| \|_    _| \|_   |
 \|___|            \|__| FAX: (716)271-3043        \|___|     \|__|     \|__|