[R-sig-eco] Question on unexpected regression results

[Rich Shepard]

On Wed, 30 Jul 2014, Bruce Miller wrote:

> I have valid ranges for the known species FA (forearm measurements) and
> Fc(minimum) and Fc (maximum). So I do two separate runs with the data
> using the lm model one with FA~Fcmin and one FA~Fcmax.
>
> The goal is to provide the predicted (estimated) values for the species
> with known FA values but w/o verified Fc value ranges.

> My concern is that the predicted values returned are much lower than the
> true values for the verified species.  Therefore I am not confident the
> predicted values for those w/o verified Fc ranges are useful.

Bruce,

   Since linear regression provides interscts and slopes for the _mean_ value
of the response variable across the range of explanatory variables, your
conundrum might be resolved by using quantile regression (R package
quantreg).

   Quantile regression looks at the relationship of the explantory
variable(s) at specified (your choice) quantiles of the response variable.
It might well be that the relationshop of Fc to FA is quite different at the
5% Fc quantile than it is at the 50% or 95% quantile.

   A starting point is Cade and Noon. 2003. A gentle introduction to quantile
regression for ecologists. Frontiers in Ecology and Environment 1:412--420.

HTH,

Rich

-- 
Richard B. Shepard, Ph.D.
Applied Ecosystem Services, Inc. | Troutdale, OR 97060 USA
www.appl-ecosys.com      Voice: 503-667-4517         Fax: 503-667-8863