I've been tweeting a lot about the role of merit in academia. A lot of readers assumed I must be thinking about race issues. This is actually not the case. I've been rethinking academic gatekeeping and how it relates to data science and statistics. 1/🧵
Many commenters saw a black man talking about "merit" and just assumed it must be about affirmative action. I care about ethnic diversity in STEM but I also think academia is too quick to gatekeep all kinds of people with non-traditional backgrounds. 2/
Imagine a world where we only allow people with the "best qualifications" to code. Our world would be much poorer for it. Yet, many research environments are highly selective in exactly this way. 3/
I think academics like me falsely assume that people can't do any useful work without lots of qualifications. So, I've been thinking that maybe the best question is "Who can best contribute to the team?" not "Who has individually accomplished the most in their life so far?". 4/
I have a lot of concerns about the reliability of data analysis. I worry that bad analyses undermine the public's faith in statistics. Bad data analysis gives the false impression that statistics is less effective than it is. 5/
Physicians that aren't really physicians can do a lot of damage to people's trust in medicine. Similarly people, who didn't train as statisticians and don't hold themselves to the same standards as the statistical community, can do a lot of damage to trust in statistics. 6/
So historically, I've been very hesitant to support everybody being involved in statistics regardless of qualifications, but I've changed my mind on this. 7/
I think we need to build a community of diverse data analysts (statisticians, data analysts, machine learning researchers, etc) who can all talk to each other and all feel like they are part of the same common project of making data analysis more reliable. 8/
If more people understood statistics, it would be more obvious to the public what kinds of statistics were good and what kinds were bad. This would be a great outcome for those of us that care deeply about good data analysis. 🧵/
Addendum 1: I'm sure many of the same people that were pro merit-based selection and against affirmative will be anti merit-based selection if it means not automatically filtering out people without PhDs. Oh well. I accept the infinite puzzle that is the human mind.
