Specifically, the R package standard, with additional R authoring tools, provides a robust framework for organizing and sharing reproducible research products. Instead of developing another standard, we suggest borrowing existing standards and practices from software engineering. Standardization of protocols for sharing would be beneficial but, such standards have not yet emerged.
Further, the practical value of sharing is often limited by poor documentation, incompatible file formats, and lack of organization, resulting in low rates of reproducibility (Hardwicke et al. 2018). Although more transparent communication of these research products has recently been encouraged (Houtkoop et al. 2018 Lindsay 2017 Vanpaemel et al. 2015 Wicherts et al. 2006 Klein et al. 2018 Martone, Garcia-Castro, and VandenBos 2018 Rouder, Haaf, and Snyder 2019 Rouder 2016), these efforts often focus narrowly on sharing data (and sometimes analysis code). However, the potential added value of these products is not fully realized due to limited sharing and curating practices. Research projects produce experiments, data, analyses, manuscripts, posters, slides, stimuli and materials, computational models, and more.
0 kommentar(er)
