On 30 Nov 1999 07:23:19 -0800, [EMAIL PROTECTED] (Richard
M. Barton) wrote:
> ...
> Scenario:
>
> Ben has a number of 1 meter square plots where he placed one or more seeds:
> 50 plots with 1 seed
> 10 plots with 25 seeds
> 10 plots with 50 seeds
>
> He replicated that design with four species of seeds.
- once you figure out what the parameters are, then I suppose that
you can compare them. Put off that question until last.
> He visited the plots every day for a week to count the number of seeds remaining.
>
>
> So the questions of interest are:
> a) Does density have an effect on seed survival?
> b) Does species have an effect on survival?
> c) What does the data look like over time?
>
> We considered modelling/analyzing the data in two ways (using SAS):
>
> 1) with seed as the unit of analysis, using Proc Lifetest to generate survival
>curves.
> The problem: for medium and high density plots, seeds would not seem to be
> independent.
- In (1), what do you mean by "the problem"? - that is the substance
of (a), isn't it?
- As in (c), I wonder, What *does* something look like over time,
and isn't that an over-complication? If you don't have 50% loss in
any of the groups, then you will have little loss of power and
generality if you compare the final counts, and not the curves, for
almost all of your testing.
- If the curves do matter across time, then you need to worry more
about How and WHY they might differ. And, what is the nature of
comparison that you actually seek between species: Is it mere
"difference" or do you have in mind some aspect of "better"?
> 2) with plot as the unit of analysis, using GLM to get a mixed model, where
> time is a repeated measure and density and species are between groups factors.
> Problem: low density (1 seed) plots have a dichotomous outcome, so much
> of the data is non-normal.
- Is there any reason not to treat the 50 plots with 1 seed as if it
were "1 plot with 50 seeds"? - to be compared to the other 10 plots
with 50 seeds. And 10 with 25 seeds. Off hand, I don't think of any
reason.
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
