Published December 15, 2021 | Version 1
Presentation Open

Do science, for software's sake!

  • 1. UT Austin

Description

In a talk given to the Workshop on the Science of Scientific-Software Development and Use (12-15 Dec 2021; virtual), I present some results from a study of scientific software development with James Howison and Caifan Du. I provide a narrative from our data and our taxonomies of 1) grant-funded scientific software development organizations and 2) transitions that those organizations can make to pursue sustainability for their software. The goal of this talk is to give audience members (many of whom were unfamiliar with this style of research) a description of how data collection and analysis proceeded as well as an argument for the value of this kind of research. I relate the narrative and taxonomies to theoretical concepts from sociological literature to show invisible work in scientific peer production. I mention ongoing efforts and relevant literature to show that the visibility of scientific software work is changing but that the audience has opportunity to shape that process for their benefit. By establishing desired metrics and working with scholars of science, the scientific software community can directly affect the research conducted about their own work. Social studies of science can both inform software work and, as I have demonstrated, help argue for its valuation.

Initial study results presented here have been published in prior workshop papers; journal article is in progress.

Talk script is in slide notes. All illustrations created by the author (contact author if you'd like individual images or Illustrator files for your own use).

A custom font was used, Cooper Hewitt from Chester Jenkins. Without it, slides will likely be weirdly formatted and kind of ugly on your own computer.

References in final slide.

Files

Files (15.5 MB)

Name Size Download all
md5:17bd8e8a0f0c283e5ec771361eabefef
15.5 MB Download