PNAS – Proceedings of the National Academy of Sciences of the United States of America
[Accessed 15 Feb 2020]
Veridical data science
Bin Yu and Karl Kumbier
PNAS first published February 13, 2020. https://doi.org/10.1073/pnas.1901326117
Predictability, computability, and stability (PCS) are three core principles of data science. They embed the scientific principles of prediction and replication in data-driven decision making while recognizing the central role of computation. Based on these principles, we propose the PCS framework, including workflow and documentation (in R Markdown or Jupyter Notebook). The PCS framework aims at responsible, reliable, reproducible, and transparent analysis across fields of science, social science, engineering, business, and government. It can be used as a recommendation system for scientific hypothesis generation and experimental design. In particular, we propose (basic) PCS inference for reliability measures on data results, extending statistical inference to a much broader scope as current data science practice entails.
Building and expanding on principles of statistics, machine learning, and scientific inquiry, we propose the predictability, computability, and stability (PCS) framework for veridical data science. Our framework, composed of both a workflow and documentation, aims to provide responsible, reliable, reproducible, and transparent results across the data science life cycle. The PCS workflow uses predictability as a reality check and considers the importance of computation in data collection/storage and algorithm design. It augments predictability and computability with an overarching stability principle. Stability expands on statistical uncertainty considerations to assess how human judgment calls impact data results through data and model/algorithm perturbations. As part of the PCS workflow, we develop PCS inference procedures, namely PCS perturbation intervals and PCS hypothesis testing, to investigate the stability of data results relative to problem formulation, data cleaning, modeling decisions, and interpretations. We illustrate PCS inference through neuroscience and genomics projects of our own and others. Moreover, we demonstrate its favorable performance over existing methods in terms of receiver operating characteristic (ROC) curves in high-dimensional, sparse linear model simulations, including a wide range of misspecified models. Finally, we propose PCS documentation based on R Markdown or Jupyter Notebook, with publicly available, reproducible codes and narratives to back up human choices made throughout an analysis. The PCS workflow and documentation are demonstrated in a genomics case study available on Zenodo.