Robustness to Failure of Assumptions of Tests for a Common Slope Amongst Several Allometric Lines – A Simulation Study

School of Mathematics and Statistics, University of New South Wales, NSW 2052, Australia.
Biometrical Journal (Impact Factor: 0.95). 04/2007; 49(2):286-99. DOI: 10.1002/bimj.200510263
Source: PubMed


In allometry, researchers are commonly interested in estimating the slope of the major axis or standardized major axis (methods of bivariate line fitting related to principal components analysis). This study considers the robustness of two tests for a common slope amongst several axes. It is of particular interest to measure the robustness of these tests to slight violations of assumptions that may not be readily detected in sample datasets. Type I error is estimated in simulations of data generated with varying levels of nonnormality, heteroscedasticity and nonlinearity. The assumption failures introduced in simulations were difficult to detect in a moderately sized dataset, with an expert panel only able to correct detect assumption violations 34–45% of the time.
While the common slope tests were robust to nonnormal and heteroscedastic errors from the line, Type I error was inflated if the two variables were related in a slightly nonlinear fashion. Similar results were also observed for the linear regression case. The common slope tests were more liberal when the simulated data had greater nonlinearity, and this effect was more evident when the underlying distribution had longer tails than the normal. This result raises concerns for common slopes testing, as slight nonlinearities such as those in simulations are often undetectable in moderately sized datasets. Consequently, practitioners should take care in checking for nonlinearity and interpreting the results of a test for common slope. This work has implications for the robustness of inference in linear models in general. (© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

Download full-text


Available from: David I Warton
  • Source
    • "ANCOVA was therefore used to test the main temporal effects of the factors, while controlling for the effects of the covariate (depth). Indeed, ANCOVA is commonly accepted to be robust to non-normality, heteroscedasticity and most failure of assumptions (Olejnik and Algina, 1984; Warton, 2007). All statistical analyses were implemented within the R programming environment (R Development Core "
    [Show abstract] [Hide abstract]
    ABSTRACT: In coastal estuarine embayments, retention of water masses due to coastal topography may result in an increased contribution of continental organic matter in food webs. However, in megatidal embayments, the effect of topography can be counterbalanced by the process of tidal mixing. Large amounts of continental organic matter are exported each year by rivers to the oceans. The fate of terrestrial organic matter in food webs of coastal areas and on neighboring coastal benthic communities was therefore evaluated, at multi-trophic levels, from primary producers to primary consumers and predators. Two coastal areas of the French Atlantic coast, differing in the contributions from their watershed, tidal range and aperture degree, were compared using carbon and nitrogen stable isotopes (δ13C and δ15N) during two contrasted periods. The Bay of Vilaine receives large inputs of freshwater from the Vilaine River, displaying 15N enriched and 13C depleted benthic communities, emphasizing the important role played by allochtonous inputs and anthropogenic impact on terrestrial organic matter in the food web. In contrast, the Bay of Brest which is largely affected by tidal mixing, showed a lack of agreement between isotopic gradients displayed by suspended particulate organic matter (SPOM) and suspension-feeders. Discrepancy between SPOM and suspension-feeders is not surprising due to differences in isotopes integration times. We suggest further that such a discrepancy may result from water replenishment due to coastal inputs, nutrient depletion by phytoplankton production, as well as efficient selection of highly nutritive phytoplanktonic particles by primary consumers.
    Full-text · Article · Dec 2013 · Estuarine Coastal and Shelf Science
  • Source
    • "The number of iterations was set to 10,000 and the critical p-value to 0.05. We calculated the SMA residuals ( y À bx) and tested the assumption of normally distributed residuals (Warton, 2006) with a Shapiro–Wilks test in Statistica 5.0 (Statsoft, 1999). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Patterns of reproductive allocation are expected to differ between species according to temporally and spatially variable costs of reproduction. Even when reproductive allocation patterns are the same, species can also differ in how the reproductive effort is allocated between offspring number and size. In this study, we compared the reproductive allocation patterns and the offspring number vs. size trade-off in two frog species, Physalaemus biligonigerus and P. gracilis, using bivariate (standardized major axis) and multiple linear regressions. Both species showed a common slope between body size and reproductive effort and thus a similar allocation pattern although P. biligonigerus has a larger body size (shift along common slope) and makes a lower reproductive effort (shift in intercept) than P. gracilis. We suggest that similar allocation patterns may be related to the shared phenologies of these frogs and that the differences in reproductive effort could represent either an adaptive shift (e.g., change in body space for the clutch) or a historical constraint. There was a negative correlation between fecundity and egg size in P. biligonigerus but not in P. gracilis as predicted by the acquisition–allocation model (Y-model). This study constitutes the first valid test of the Y-model based on recent predictions derived for the trade-off between offspring size vs. number. We conclude that future studies should compare reproductive allocation patterns between species using tests of allometric slopes with appropriate phylogenetic control to detect both adaptive shifts in allocation strategies and correlations with other life-history traits.
    Full-text · Article · Jul 2008 · Acta Oecologica
  • Source

    Preview · Article ·
Show more