Dear all,

This is my maidenpost (I am a new member since 5 minutes) to this list. My
problem is the following: I have a dataset with morphometric data from
fishes (33 measurements per fish; 11 males and 10 females per sample
location; 3 different sample locations; 2 different fish species).

I want to perform a PCA on these data, but I am wondering wether I should
use the real data (log transformed) or wether I should use the residuals
for the PCAnalysis?

(After having done the PCA, I want to select the most contributing
variables through a Factor Analysis; then I will do an ANOVA on each of the
selected variables to investigate which samples are different from each
other)

Kind regards,
Dieter Anseeuw
[EMAIL PROTECTED]

Reply via email to