<p>Data Curation and the Data Science Workflow</p>

This note proposes and illustrates an extended version of the data science "whole game" offered by Wickham, Çetinkaya-Rundel, and Grolemund (2023). The extended version divides the data science whole game into two processes: Curate and Understand. Using the case of Australian end-of-day stock price data, I explain what the Curate process is and how it covers important elements of the data science whole game that are neglected in the original version. One feature of Curate is that it requires a set of specialist skills often not possessed by practitioners expert in the Understand phase. As such, I argue that a more effective division of labour might result from better delineating the two data science processes.