[rmc]

It can tell you why people left. Which might tell you the problems in it.

Update: Imagine a company has many people leaving, and they ask why. If everyone says "Got paid better elsewhere" it means you have a problem with salary. If "Sick of commuting out to the sticks", maybe look at relocation. If "Tech is old fashioned and boring", maybe look at what your employees are working on. If "Manager is an asshole", maybe fire/retrain that manager. etc.

[malandrew]

So for the 716 people who left, every single one of them only had sexist reasons for leaving? None of them found out the work wasn't that appealing? None had family issues that forced them to leave to care for a sick or ailing parent? None had a medical issue of their own that meant they could no longer work? None simply found something else they were more passionate about? None simply earned their FU money at a successful company and decided maybe they want to dedicate themselves to philanthropic pursuits or just lounging around? None decided that they prefer a job with more social interaction over one where you spend many hours at a desk solving problems quietly?

All the above are reasons that I've heard as reasons for people leaving our industry and other industries. They apply to both men and women. Where are these other reasons in the conclusions of this study?

Where are the opinions of women like Susan Sons and Meredith Patterson who quite like nerd culture and feel it's misunderstood by many of those writing about what's wrong with tech culture? [0]. This study is so incredibly one-sided in its conclusions that I can only conclude that it went out of its way to cherry-pick the study participants. Or its possible that it was subject to an unconscious bias of its own. Perhaps the author is so thoroughly in her own echo chamber with regard to these issues that when she reached out through her network to find people willing to talk to her about why they left, she ended up recruiting largely from a population that is biased towards confirming the conclusions she wanted to demonstrate because that's what her social network experiences disproportionately. When examining the biases of others, it's important to examine your own and not commit the same mistakes as those you are are trying to correct. Any "study" without a discussion of sampling methodology to control for biases and identifying those biases is typically not constructive.

[0] https://medium.com/@maradydd/when-nerds-collide-31895b01e68c