Distance‐based multivariate analyses confound location and dispersion effects

By David I. Warton, Stephen T. Wright, Yi Alice Wang. 2012

1,114 citations recorded by the source corpus.

Subfields: Ecological Modeling, Environmental Chemistry, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/0b512924e81bee13269dca20b594ce843d315902

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 Environmental Chemistry

In Nature and Landscape Conservation

Citations in the corpus

mvabund– an R package for model‐based analysis of multivariate abundance data (2012, background): “By default, the multivariate test statistic is calculated assuming independence of species response variables, which makes the test statistic a simple sum of the univariate test statistics across species, as in Warton et al. (2012).”
Biodiversity consequences of land-use change and forest disturbance in the Amazon:a multi-scale assessment using ant communities (2016, methodology): “As both methods yielded very similar results, we opted to show only those from PERMANOVA.”
Comparing colors using visual models (2018, background): “Like other distance-based methods, it is sensitive to confounding variance heterogeneity among samples when testing for differences in location (Warton et al. 2012; Anderson and Walsh 2013).”
The fourth‐corner solution – using predictive models to understand how species traits interact with the environment (2014, methodology): “There has been a recent trend towards the development of model-based approaches to analysing multi-species data (Dunstan et al. 2011; Ives & Helmus 2011; Ovaskainen & Soininen 2011; Warton et al. 2012) and in this paper we have showed that such a method can be used to solve the “fourth corner proble”
Modelling Palaeoecological Time Series Using Generalised Additive Models (2018, methodology): “One currently available avenue for fitting a multivariate GAM is via regularised sandwich estimators and GLMs (Warton, 2011), which involves fitting separate GLMs (or GAMs) to each response variable and subsequently using resamplingbased hypothesis tests to determine which covariates are related to ”
Using metabarcoding to ask if easily collected soil and leaf-litter samples can be used as a general biodiversity indicator (2014, methodology): “…predictors on community composition. mvabund is a multivariate implementation of generalized linear models, and, unlike dissimilarity-matrix-based methods, mvabund does not confound location with dispersion effects, which can inflate type 1 and 2 errors (Wang et al., 2012; Warton et al., 2011).”
So Many Variables: Joint Modeling in Community Ecology. (2015, methodology): “For example, models can be used to account for important (and otherwise spurious) data properties such as the mean–variance relationship [56].”
Reliable, verifiable and efficient monitoring of biodiversity via metabarcoding. (2013, background): “This disagreement could result from suboptimal solution finding in the NMDS ordination and/or from the community-level heteroscedasticity that causes Type 1 and 2 errors in dissimilarity-matrix-based analyses (Warton et al. 2012).”
Compositional analysis: a valid approach to analyze microbiome high-throughput sequencing data. (2016, background): “Unfortunately, distance-based tools can confuse location (difference) and dispersion (variance) effects (Warton et al. 2012), and so additional approaches based on a compositional paradigm should be used for exploratory data analysis.”