Ana Trisovic
[introductory/advanced] Principles, Statistical and Computational Tools for Reproducible Data Science
Summary
Computational reproducibility is the ability to replicate or reproduce the results of an experiment or study using the same methods, data, and analysis as the original study. It is a crucial aspect of data science research, as it allows researchers to confirm the validity and reliability of results and conclusions and to build on previous research. This course is designed to provide students with the knowledge and skills necessary to conduct research in a reproducible and transparent manner. It covers various aspects of reproducible research, including experiment design, data lifecycle from data collection to data sharing, FAIR principles, data provenance, statistical methods and computational tools such as workflows, version control, and containerization. The course will use practical examples in statistical programming languages R and Python.
The course is suitable for anyone doing data-intensive research across various disciplines, and it offers a comprehensive overview of reproducible research, best practices, and big data management.
Upon completion of this course, students will have the skills to conduct research in a reproducible and transparent manner and be equipped to tackle the challenges of reproducibility in their future research endeavors.
Syllabus
- Fundamentals of reproducible research
- Experiment design and data lifecycle
- Data provenance, workflows and computational tools
- Statistical methods for reproducible research
References
National Academies of Sciences, Engineering, and Medicine. Reproducibility and replicability in science. National Academies Press, 2019.
Arnold, Becky, Louise Bowler, Sarah Gibson, Patricia Herterich, Rosie Higman, Anna Krystalli, Alexander Morley, Martin O’Reilly, and Kirstie Whitaker. The Turing Way: a handbook for reproducible data science. Zenodo, 2019.
Trisovic, Ana, Matthew K. Lau, Thomas Pasquier, and Mercè Crosas. “A large-scale study on research code quality and execution.” Nature Scientific Data.
Curtis Huttenhower, John Quackenbush, Lorenzo Trippa, Christine Choirat. Principles, Statistical and Computational Tools for Reproducible Data Science. edX.
Pre-requisites
Basic knowledge of statistics, programming and command line.
Short bio
Ana Trisovic is a Research Associate at Harvard University. Her work focuses on research reproducibility, big data workflows, machine learning, data repositories and climate, geospatial and health data engineering. Before coming to Harvard, she was a postdoctoral fellow at the University of Chicago. She completed her Ph.D. in computer science from the University of Cambridge and CERN in 2018.
