Does it really say that or are women just slightly more probable to enter the field without a degree?

And I'd argue it's a pretty bad proxy. Because the field might be growing (or shrinking) and percentages don't mean anything. 23% of 10k is less than 20% of 5k, for example. The percentage numbers don't really indicate whether someone will stay in the field, it's just a number that's highly dependent on a lot of variables and a very bad indicator for "people are staying in the field". I'm happy to be corrected, it's just how I read this.

Additionally, if your assumption is that 23%>20%, that would kind of mean that it's capped at 23%, right? Once more the CS degree quota is higher than 23%, following your logic, that would be an indicator that women are more likely to leave the field because it naturally gravitates towards 23%. But that's not based on anything, you could argue just as well that it's an indicator that more women are starting to take interest in CS as a career.

> they're more likely to remain in the field if they are a woman

Top earning fields (+most fields) were rife with strong resistance to hiring women. For women who'd managed jobs in top-earning professions (<pay) - this was constant, persuasive pressure to stay where they were.

source: grew up around professional women born early 1920s (budget analyst, peace corps, navy intel, usvp sec).

consider that a lot of the culture in tech is also there for the first four years of undergrad, and so 23% often represents the people who basically made it through four years. are people who have experienced it for four years likelier to put up with more of the same?