Date:         Mon, 1 Feb 2010 11:14:11 -0800
Reply-To:     tarpon <tarponquest@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         tarpon <tarponquest@GMAIL.COM>
Organization: http://groups.google.com
Subject:      nested design ?'s
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

hey guys:
i collected larval fish at a few different sampling sites a few hours
apart--morning and afternoon-- (~ 4 hours) for a our month period to
determine if the abundance measurements changed during the day. i have
written the following code:
of course, i'm willing to listen to crictism of code and what
not......thanks in advance for an input

proc glm data=larval_fish;
class tod_id site_id julian_date ;
model density= site_id julian_date(site_id) tod_id
(site_id*julian_date);
means tod_id(site_id*julian_date);
run;
quit;

where site_id = specific sampling site
tod_id= morning or afternoon sample
julian date=1 to 365 (starting with january 1st as "1" and december
31st as "365"---for those of you who haven't see this before....)


my output.....i'm trying to determine the proper df for the
interactions (see below) to report...i understand that densities at
sampling sites are statistically dissimilar (F=9.8, p =<.0001), that
density at a specific site is dissimliar between sample dates (julian
date) (F=3.79, p= <.0001) and the last interaction, and most important
to me, that a comparsion of fish density between the morning and
afternoon at a specific sample site on the same date is dissimliar
(F=2.1, p = .0014)


                                                       The GLM
Procedure

Dependent Variable: density

                                                              Sum of
                      Source                      DF
Squares     Mean Square    F Value    Pr > F

                      Model                       74
72.8700523       0.9847304       3.92    <.0001

                      Error                      121
30.3659822       0.2509585

                      Corrected Total            195     103.2360345


                                     R-Square     Coeff Var      Root
MSE    density Mean

                                     0.705859      123.7053
0.500958        0.404960


                      Source                      DF       Type I
SS     Mean Square    F Value    Pr > F

                      site_id                     10
26.91048951      2.69104895      10.72    <.0001
                      julian_date(site_id)        27
26.48996349      0.98110976       3.91    <.0001
                      tod_i(site_i*julian)        37
19.46959932      0.52620539       2.10    0.0014


                      Source                      DF     Type III
SS     Mean Square    F Value    Pr > F

                      site_id                     10
24.58203525      2.45820352       9.80    <.0001
                      julian_date(site_id)        27
25.64715356      0.94989458       3.79    <.0001
                      tod_i(site_i*julian)        37
19.46959932      0.52620539       2.10    0.0014