Transparency, usability, and reproducibility: Guiding principles for improving comparative databases using primates as examples

Abstract

This is the accepted version of the following article: Borries, C., Sandel, A. A., Koenig, A., Fernandez-Duque, E., Kamilar, J. M., Amoroso, C. R., Barton, R. A., Bray, J., Di Fiore, A., Gilby, I. C., Gordon, A. D., Mundry, R., Port, M., Powell, L. E., Pusey, A. E., Spriggs, A. and Nunn, C. L. (2016), Transparency, usability, and reproducibility: Guiding principles for improving comparative databases using primates as examples. Evolutionary Anthropology: Issues, News, and Reviews, 25(5): 232-238, which has been published in nal form at https://doi.org/10.1002/evan.21502. This article may be used for non-commercial purposes in accordance With Wiley Terms and Conditions for self-archiving.Additional information: Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-pro t purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. From the beginning of evolutionary biology, the comparative method has been a major analytical tool, 1-3 allowing for the examination of patterns and processes of evolutionary change. 4 Some of the main obstacles to overcome in comparative analyses have been statistical in nature:How should we control for confounding variables? What criteria should we use to assess whether patterns are statistically significant and biologically meaningful? How should we control for the non-independence of comparative data that stems from phylogenetic relatedness? Much progress has been made with respect to these issues, especially in the development and use of 3 phylogenetic comparative methods. 3,5,6 For example, building on initial descriptions of phylogenetically independent contrasts, 7 methods can now incorporate phylogenetic uncertainty, 8,9 intraspecific variation, 10-12 and different models of phenotypic evolution. 13,14 Although phylogenetic and statistical methods are rapidly advancing, an increasing number of researchers argue that the data to which these methods are applied are 'stuck in the dark ages'. [15][16][17] It is imperative that, before the specific methods employed in a comparative study are considered, the suitability of the data be thoroughly evaluated. The time has come to bring our comparative databases into the modern age, and to represent uncertainty in the data in the same way we might represent uncertainty in a statistical model or in a phylogeny. 18 It is also important that we be able to evaluate which sources of uncertainty -in the data, the phylogeny, and the statistical methods -have the greatest influence on comparative results.To approach these issues, the authors met on May 28, 2014 at the National Evolutionary Synthesis Center (NESCent, Durham, NC,...

Comparative analyses of basal rate of metabolism in mammals: data selection does matter

“…Our analyses substantiate the arguments marshalled by Borries et al . (), who emphazised the need for more careful reliability control with data used in comparative analyses. These authors strongly recommended more transparency in comparative databases, including a better description of the methods applied in the original papers.…”

“…For birds, McKechnie & Wolf () have shown that different scaling equations for BMR may be obtained depending on the severity of data selection. More generally, databases including descriptions and/or evaluations of the conditions under which the original data were obtained have been emphatically recommended (Borries et al, ).…”

The evolution of pair‐living, sexual monogamy, and cooperative infant care: Insights from research on wild owl monkeys, titis, sakis, and tamarins

“…More recently, Borries, Gordon, and Koenig ()—concerned about the differences they found between their own literature‐based assessments of gestation length estimates for various Asian colobines and macaques and those published in four “popular” life history compilations—attempted to reproduce the analyses in those compilations without success. Some of the problems they found, as well as other issues related to the transparency, usability and reproducibility of comparative analyses are discussed in more detail in a recent publication (Borries et al, ).…”

“…For example, in making our own estimates of infanticide risk below, we use both the minimum and maximum reported values for age at weaning; the former “favors” the hypothesis, while the latter “counters” it. We also suggest that, given the paucity of data, authors must make clear the definitions and criteria that are used for assigning data to a taxon, providing explicit and accurate reference to the primary literature where the original data were published (Borries et al, ).…”

“…The comparative study of “monogamy” is no exception to our proposition that it requires judicious compilation and integration of data from a large number and a wide range of sources that vary in quality, length, scope, and sample size (Borries et al, ). The goal should be to produce research that: (a) focuses on the natural history of the taxa of interest; (b) uses methods that are either standardized, or comparable, across research sites and taxa; and (c) yields data on the variables of interest and on variation in those traits.…”