Sarah Emerson, Charlotte Wickham, and Kenneth J. Ruzicka Jr. 2015. Rethinking the linear regression model for spatial ecological data: Comment. Ecology 96:17311735. http://dx.doi.org/10.1890/14-0879.1


Supplement

R code used to perform simulations.
Ecological Archives E096-178-S1.

Copyright


Authors
File list (downloads)
Description


Author(s)

Sarah Emerson
Department of Statistics, Oregon State University
44 Kidder Hall
Corvallis, OR 97331-4606
E-mail: emersosa@stat.oregonstate.edu

Charlotte Wickham
Department of Statistics, Oregon State University
44 Kidder Hall
Corvallis, OR 97331-4606
E-mail: charlotte.wickham@stat.oregonstate.edu

Kenneth J. Ruzicka Jr.
Department of Forest Ecosystems and Society, Oregon State University
321 Richardson Hall
Corvallis, OR 97331-4606
Kenneth.ruzicka@oregonstate.edu


File list

simulations.R (MD5: bdda5503ab6ec0d1374d340d60f562d6)

Description

This comment used the attached R script to conduct simulation studies of spatial component regression (SCR). The file simulations.R contain all code needed to run the simulations to test SCR performance for three objectives: (1) inference under the null hypothesis; (2) inference when the predictor of inference does have an effect on the outcome; and (3) matrix selections. The code will simulate 16 sets of 1000 data sets each. The 16 data sets represent all possible combinations of 2 different spatial predictor types 4 autocorrelation types and 2 effect sizes for the spatial predictor.