Recent decades have seen rapid development of new analytical methods to investigate patterns of interspecific variation. Yet these cutting-edge statistical analyses often rely on data of questionable origin, varying accuracy, and weak comparability, which seem to have reduced the reproducibility of studies. It is time to improve the transparency of comparative data while also making these improved data more widely available. We, the authors, met to discuss how transparency, usability, and reproducibility of comparative data can best be achieved. We propose four guiding principles: 1) data identification with explicit operational definitions and complete descriptions of methods; 2) inclusion of metadata that capture key characteristics of the data, such as sample size, geographic coordinates, and nutrient availability (for example, captive versus wild animals); 3) documentation of the original reference for each datum; and 4) facilitation of effective interactions with the data via user friendly and transparent interfaces. We urge reviewers, editors, publishers, database developers and users, funding agencies, researchers publishing their primary data, and those performing comparative analyses to embrace these standards to increase the transparency, usability, and reproducibility of comparative studies.
|Número de páginas
|Evolutionary Anthropology: Issues, News, and Reviews
|Publicada - 1 sep. 2016
|Publicado de forma externa