Year: 2014
Type: article
Abstract: One of the hallmarks of scientific progress is the ability to identify and control or reduce newly discovered sources of bias and error. In response to emerging evidence that a great deal of scientific work may not be reproducible, the National Institutes of Health recently launched a reproducibility initiative to validate data published in landmark studies. The editors of Nature and Science have also announced steps that their journals will take to promote reproducibility. Efforts by scientific groups to promote reproducibility are focused mostly on measures aimed at reducing biases and errors related to experimental design and statistical analysis, some of which include reporting of data and methods fully, using control groups and blinding, and ensuring that studies are adequately powered. Although implementing such measures is essential to promoting reproducibility, it is also important for scientists to address bias and errors affecting their own judgment and reasoning. Taking human error seriously has significant ramifications for scientific practice, communication, and policy.
Source: BioScience
Authors: Kevin C. Elliott, David B. Resnik
Language: English
Subfield: Information Systems and Management
Field: Decision Sciences
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W2123852619