Model selection and model averaging in behavioural ecology: the utility of the IT-AIC framework

By S. Richards, Mark J. Whittingham, Philip Stephens. Behavioral Ecology and Sociobiology, 2010

522 citations recorded by the source corpus.

Behavioural ecologists often study complex systems in which multiple hypotheses could be proposed to explain observed phenomena. For some systems, simple controlled experiments can be employed to reveal part of the complexity; often, however, observational studies that incorporate a multitude of causal factors may be the only (or preferred) avenue of study. We assess the value of recently advocated approaches to inference in both contexts. Specifically, we examine the use of information theoretic (IT) model selection using Akaike’s information criterion (AIC). We find that, for simple analyses, the advantages of switching to an IT-AIC approach are likely to be slight, especially given recent emphasis on biological rather than statistical significance. By contrast, the model selection approach embodied by IT approaches offers significant advantages when applied to problems of more complex causality. Model averaging is an intuitively appealing extension to model selection. However, we were unable to demonstrate consistent improvements in prediction accuracy when using model averaging with IT-AIC; our equivocal results suggest that more research is needed on its utility. We illustrate our arguments with worked examples from behavioural experiments.

Subfields: Ecological Modeling, Ecology, Evolution, Behavior and Systematics, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/367657106fbeb03afbdf9a55fccbf00bc6a052d5

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Ecological Modeling

In Ecology, Evolution, Behavior and Systematics

In Nature and Landscape Conservation

Citations in the corpus

A farewell to the sum of Akaike weights: The benefits of alternative metrics for variable importance estimations in model selection (2017, background): “It has also been shown to be less severe for full model-averaged metrics than when considering variable importance metrics in the first ranked model only (Burnham & Anderson, 2002; pp. 151–153, but see Richards, Whittingham, & Stephens, 2011).”
Using information theory as a substitute for stepwise regression in ecology and behavior (2010, unknown): “…estimates of the main effects towards their values in models without interactions, and (2) individual main effect parameter estimates themselves may often be biased in the presence of interactions (Aiken and West 1991; see Richards et al. 2010 for further concerns regarding model averaging).”
Issues in information theory-based statistical inference—a commentary from a frequentist’s perspective (2010, methodology): “…approach to statistical inference Since approximately a decade, a new approach to statistical inference has become increasingly popular in applied statistics used in research in ecology and, to a lesser extent, also animal behaviour (Garamszegi et al. 2009; Garamszegi 2010; Richards et al. 2010).”
A brief guide to model selection, multimodel inference and model averaging in behavioural ecology using Akaike’s information criterion (2010, methodology): “applied to analyses of mechanistic models of behaviour and controlled experiments (see, e.g. Lau and Glimcher 2005; Luttbeg et al. 2009; Richards et al. 2010), where the problems of data dredging are mitigated.”
Model averaging, missing data and multiple imputation: a case study for behavioural ecology (2010, methodology): “As a result, model averaging offers more reliable and robust point and uncertainty estimation of parameters (for more detailed descriptions model averaging, see Burnham and Anderson 2002; Richards et al. 2010; Symonds and Moussalli 2010).”
Multimodel inference in ecology and evolution: challenges and solutions (2011, background): “…U T I O N A R Y B I O L O G Y ª 2 0 1 1 E U R O P E A N S O C I E T Y F O R E V O L U T I O N A R Y B I O L O G Y stable than those based on choosing the best model, as the former is less likely to erroneously conclude that weak parameter estimates are zero (Richards, 2005; Richards et al., 2010).”
Early assessment of the impact of alien species: differential consequences of an invasive crayfish on adult and larval amphibians (2011, methodology): “Per each model i, we calculated the AIC weight wi, which is the probability for a model to be the best one among the candidates (Richards et al., 2011).”