February 19, 2021 – Russ Poldrack

Russ Poldrack, Professor of Psychology. Stanford University

Toward a culture of computational reproducibility: Values, principles, and practices

Ensuring that the results of data analysis are both valid and reproducible is a fundamental responsibility of every computational scientist, but both are increasingly difficult in the context of complex analysis workflows and big data.  Building off of ideas from software engineering, I will argue that we need to embrace a culture of computational reproducibility.  I will outline a set of values that motivate this work and principles that guide the work, and then focus on a set of practices that can help improve reproducibility in computational science. I will conclude by addressing some potential concerns about the impacts of this cultural shift.