I question the credibility of this article. There are factors missing in this analysis, education for instance. I am also weary of an article based on `self-reported interest.`
Out of curiosity, how would one measure interest without self-reporting? Constant surveillance to make sure women aren't secretly reading Dr Dobbs (RIP) in an incognito browser window while filling their history with TMZ?
You can measure interest more objectively by seeing how many start down a career path, for example. Another measure could be to look at proxies to interest - such as hobby programming as a proxy for interest in computing, or reading books as a proxy for writing.
But, those are much harder to measure than just counting "likes" on a web form.