Contrasting approaches to statistical regression in ecology and economics
Article first published online: 3 MAR 2009
© 2009 The Authors. Journal compilation © 2009 British Ecological Society
Journal of Applied Ecology
Volume 46, Issue 2, pages 265–268, April 2009
How to Cite
Armsworth, P. R., Gaston, K. J., Hanley, N. D. and Ruffell, R. J. (2009), Contrasting approaches to statistical regression in ecology and economics. Journal of Applied Ecology, 46: 265–268. doi: 10.1111/j.1365-2664.2009.01628.x
- Issue published online: 3 MAR 2009
- Article first published online: 3 MAR 2009
- Received 16 November 2008; accepted 15 January 2009Handling Editor: Marc Cadotte
- land use;
- spatial autocorrelation;
- structural equation model;
- 1Conservation and natural resource management challenges are as much social problems as biological ones. In recognition of this fact, ecologists and economists work increasingly closely together. We discuss one barrier to effective integration of the two disciplines: put simply, many ecologists and economists approach statistical regression differently.
- 2Regression techniques provide the most commonly used approach for empirical analyses of land management decisions. Researchers from each discipline attribute differing importance to a range of possibly conflicting design criteria when formulating regression analyses.
- 3Ecologists commonly attribute greater importance to spatial autocorrelation and parsimony than do economists when designing regressions. Economists often attribute greater importance than ecologists to concerns about endogeneity and conformance with a priori theoretical expectations.
- 4Synthesis and applications. The differing importance attributed to different design characteristics may reflect a process of cultural drift within each discipline. Greater interdisciplinary collaboration can counteract this process by stimulating the flow of ideas and techniques across disciplinary boundaries.