----- Forwarded message from Rodrigo Lima -----
Date: Mon, 28 May 2012 15:48:16 -0400
From: Rodrigo Lima
Reply-To: Rodrigo Lima
Subject: standardizing environmental variables
To: "[email protected]"
Dear morphometricians,
I'm running a PLS using Procrustes distances in the first block and
environmental variables, latitude, and longitude in the second block. I'm
analyzing populations with different numbers of specimens (from 5 to 50
specimens), and the environmental variables have one value for all individuals
in each population (e.g. all individuals from population 1 have the following
values: latitude 50.5, longitude -60.2, temperature 23.7, precipitation 900,
etc.). I'm doing it in PAST, and I believe it doesn't standardize the variables
automatically.
I have a question about the calculation of standardized values:
1 - Should I standardize the environmental variables using all individuals in
the sample (i.e. if I have 50 individuals in population 1 and 5 individuals in
population 2, the temperature for population 1 will be represented 50 times and
the temperature for population 2 will be represented 5 times)? Or should I take
the value for each population only once (i.e. if I have 15 populations, I'll
have 15 values for temperature, 15 for precipitation, 15 for latitude, etc.)?
I hope I didn't complicate a simple question...
Thanks!
Rodrigo
----- End forwarded message -----
