R help - Mar 2005 - Help with lme Random Factor

Hi,

I need help creating a code for a multiple BACI design (Before-After
Control-Impact) ANOVA.  I'm new to R and basically need to run a complex
mixed model ANOVA that treats location as a random factor.

Data are from a fire experiment, run 2001-2004 (2 years pre, 2 years
post).  Response is bird abundance.  4 Treatments had 3 replicates each
(forest stands): 1. Control, 2. Prescribed fire only, 3. Timber harvesting
only, and 4. Fire and harvesting.

Basically, I'm having trouble figuring out the code that makes the
variable "LOC" a random factor.

Also, I'm having trouble figuring out how to script contrasts comparing
the treatments to the control (multiple comparisons w/ a control). 
Treatment group is not a part of the lme script.  Do I need to run them
each individually, i.e. C vs. T1, C vs. T2, C vs T3? or can I code in
treatment contrasts?

Script (used S-plus GUI to build it):

GCKI~.+C+B+LOC%in%C+T%in%B+C:B+C:T%in%B+LOC%in%C:B+LOC%in%C:T%in%B
**unsure what goes here to make LOC random****

Terms:

GCKI: Golden-crowned kinglet abundance.

C:  Control-Impact (2 levels, either control (no treatment), or impact
(received treatment after 2 years)

B: Before-After (2 levels, 2001-2002 = Before, 2003-2004 = After)

T: Time, 4 levels, years 2001, 2002, 2003, 2004

LOC: location, 3 replicates in among the 4 treatments (RANDOM FACTOR)


Sorry to write so much, but any help would be appreciated.  The ANOVA
model comes from a book: Monitoring Ecological Impacts: Downes et al.
2002.

I am used to running GUIs, but it appears I need to script this ANOVA.  I
am unfamiliar with random factor scripts or scripts for multiple contrasts
with controls.

Cheers,

Andrew

Hello Andrew,

I strongly suggest that you obtain a copy of the book that documents the use of
the package, and read it.

Mixed Effects Models in S and S-Plus
by Jose C. Pinheiro and Douglas M. Bates  Springer Verlag 2000.

I hope that this helps,

Andrew


----- Original Message -----
From: Andrew James Amacher <aamacher at nature.berkeley.edu>
Date: Wednesday, March 9, 2005 4:19 pm
Subject: [R] Help with lme Random Factor
> Hi,
> 
> I need help creating a code for a multiple BACI design (Before-After
> Control-Impact) ANOVA.  I'm new to R and basically need to run a 
> complexmixed model ANOVA that treats location as a random factor.
> 
> Data are from a fire experiment, run 2001-2004 (2 years pre, 2 years
> post).  Response is bird abundance.  4 Treatments had 3 replicates 
> each(forest stands): 1. Control, 2. Prescribed fire only, 3. 
> Timber harvesting
> only, and 4. Fire and harvesting.
> 
> Basically, I'm having trouble figuring out the code that makes the
> variable "LOC" a random factor.
> 
> Also, I'm having trouble figuring out how to script contrasts 
> comparingthe treatments to the control (multiple comparisons w/ a 
> control). 
> Treatment group is not a part of the lme script.  Do I need to run 
> themeach individually, i.e. C vs. T1, C vs. T2, C vs T3? or can I 
> code in
> treatment contrasts?
> 
> Script (used S-plus GUI to build it):
> 
> GCKI~.+C+B+LOC%in%C+T%in%B+C:B+C:T%in%B+LOC%in%C:B+LOC%in%C:T%in%B
> **unsure what goes here to make LOC random****
> 
> Terms:
> 
> GCKI: Golden-crowned kinglet abundance.
> 
> C:  Control-Impact (2 levels, either control (no treatment), or impact
> (received treatment after 2 years)
> 
> B: Before-After (2 levels, 2001-2002 = Before, 2003-2004 = After)
> 
> T: Time, 4 levels, years 2001, 2002, 2003, 2004
> 
> LOC: location, 3 replicates in among the 4 treatments (RANDOM FACTOR)
> 
> 
> Sorry to write so much, but any help would be appreciated.  The ANOVA
> model comes from a book: Monitoring Ecological Impacts: Downes et al.
> 2002.
> 
> I am used to running GUIs, but it appears I need to script this 
> ANOVA.  I
> am unfamiliar with random factor scripts or scripts for multiple 
> contrastswith controls.
> 
> Cheers,
> 
> Andrew
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! http://www.R-project.org/posting-
> guide.html