“Science” of diversity
I think part of “beef” with diversity is that a lot of the computational studies of it seem to be toy problems that don’t say much and can have results flipped with different parameters. When we study diversity in the real world, I think a large part of the focus is on racial diversity (which is extremely important!), but I think the a question that gets left by the wayside is measuring diversity of thought. (perhaps because this is intractable)
I should note, the problem I have with computational diversity studies extends to any form of computational modelling where the scientist is able to build in their own biases.
Maybe you could use some sort of cosine similarity between the members of a lab as a measure of diversity. This seems dumb for multiple reasons. Firstly, people in a lab will work together, so they’ll have a lot of overlap on the kinds of papers they write. Or perhaps you encode their papers/work before they join a lab as a measure of diversity? This could be interesting and as far as metrics go, it seems as flawed/rigorous as any I’ve stumbled across.
How to avoid monoculture?
A lot of these papers point out the damage of monoculture! The most striking being the paper which showed that employers which all use an “optimal” automated hiring process can instead hiring a suboptimal subset of employees. We have strength in diversity, but there are so many pressures socially to conform.
From open questions, How can you measure tradeoff between conformity and diversity? There is strength in diversity, but it seems like there is also a lot of social push for conformity. Is there an optimal balance? What kind of metrics would someone use to measure these?
Bibliography
- Weidenmüller, Anja. “The Control of Nest Climate in Bumblebee (Bombus Terrestris) Colonies: Interindividual Variability and Self Reinforcement in Fanning Response.” Behavioral Ecology 15, no. 1 (January 2004): 120–28. https://doi.org/10.1093/beheco/arg101.
- Jones, Julia C., Mary R. Myerscough, Sonia Graham, and Benjamin P. Oldroyd. “Honey Bee Nest Thermoregulation: Diversity Promotes Stability.” Science 305, no. 5682 (July 16, 2004): 402–4. https://doi.org/10.1126/science.1096340.
- Zollman, Kevin J. S. “The Epistemic Benefit of Transient Diversity.” Erkenntnis 72, no. 1 (January 2010): 17–35. https://doi.org/10.1007/s10670-009-9194-6.
- Mastroianni, Adam. “Pop Culture Has Become an Oligopoly.” Accessed June 7, 2022. https://experimentalhistory.substack.com/p/pop-culture-has-become-an-oligopoly.
- Mayo-Wilson, Conor, Kevin J. S. Zollman, and David Danks. “The Independence Thesis: When Individual and Social Epistemology Diverge.” Philosophy of Science 78, no. 4 (October 1, 2011): 653–77. https://doi.org/10.1086/661777.
- AlShebli, Bedoor K., Talal Rahwan, and Wei Lee Woon. “The Preeminence of Ethnic Diversity in Scientific Collaboration.” Nature Communications 9, no. 1 (December 2018): 5163. https://doi.org/10.1038/s41467-018-07634-8.
- Grim, Patrick, Daniel J. Singer, Aaron Bramson, Bennett Holman, Sean McGeehan, and William J. Berger. “Diversity, Ability, and Expertise in Epistemic Communities.” Philosophy of Science 86, no. 1 (January 2019): 98–123. https://doi.org/10.1086/701070.
- Kleinberg, Jon, and Manish Raghavan. “Algorithmic Monoculture and Social Welfare.” Proceedings of the National Academy of Sciences 118, no. 22 (June 2021): e2018340118. https://doi.org/10.1073/pnas.2018340118.
- Hong, Lu, and Scott E. Page. “Groups of Diverse Problem Solvers Can Outperform Groups of High-Ability Problem Solvers.” Proceedings of the National Academy of Sciences 101, no. 46 (November 16, 2004): 16385–89. https://doi.org/10.1073/pnas.0403723101.
- Swartz, Talia H, Ann-Gel S Palermo, Sandra K Masur, and Judith A Aberg. “The Science and Value of Diversity: Closing the Gaps in Our Understanding of Inclusion and Diversity.” The Journal of Infectious Diseases 220, no. Supplement_2 (August 20, 2019): S33–41. https://doi.org/10.1093/infdis/jiz174.
- Budescu, David V., and Mia Budescu. “How to Measure Diversity When You Must.” Psychological Methods 17, no. 2 (2012): 215–27. https://doi.org/10.1037/a0027129.