Ten simple rules for initial data analysis

• Published in: PLoS computational biology Vol. 18; no. 2; p. e1009819
• Main Authors: Baillie, Mark, le Cessie, Saskia, Schmidt, Carsten Oliver, Lusa, Lara, Huebner, Marianne
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 01.02.2022; Public Library of Science (PLoS)
• Subjects: Computer and Information Sciences; Data Analysis; Data management; Data mining; Decision analysis; Humans; Hypotheses; Information management; Laws, regulations and rules; Metadata; Methods; Ovarian Neoplasms; Physical Sciences; Planning; Reproducibility; Research and Analysis Methods; Researchers; Science Policy; Social Sciences; Statistical analysis; Subject specialists; United States
• ISSN: 1553-7358; 1553-734X; 1553-7358
• DOI: 10.1371/journal.pcbi.1009819

Typically, researchers do not perform IDA in a systematic way, if at all, or mix IDA activities with subsequent data analysis tasks such as hypothesis generation or exploration, formal analysis, and interpretation of conclusions. The value of an effective IDA strategy for researchers lies in ensuring that data are of sufficient quality, that model assumptions made in the SAP are satisfied, or to support decisions for the statistical analyses (and are adequately documented). IDA requires domain knowledge, especially researchers with an understanding of why and how the data was measured and collected, expertise in data management and stewardship, competencies in planning and implementing data analysis, and experience of scientific computing practices. Make IDA reproducible IDA is a crucial part of the research pipeline, and as such, it should be well documented to promote transparency, utility, and reproducibility. [...]keeping track of changes that you and your collaborators make to project data, programs (including analysis scripts, libraries, and packages), and documentation (including plans and reports) is a key IDA practice [15].