Gender differences in Stumbleupon categories

Earlier today, Jo Brodie posted a couple of screenshots of the suggested categories presented to new users on Stumbleupon (a link-sharing website).

She pointed out that the categories were different according to the user’s indicated gender. In particular, the category “science” was only presented to male users.

I have taken the information from the screenshots and organised into the table below. I’ve lumped the categories into, er, categories, roughly reflecting the themes Stumbleupon offers to men and women.

Notice the stereotypes: men like beer, cars and babes, whereas as women are interested in celebrities, relationships and interior design.

Jo shared an update from Stumbleupon, which said that “categories are offered based on what other people of [the same] gender have selected” . On the sign-up page, it says “This helps us find the best content for you” under the gender selection question.

Jo then asks the obvious question: is it the case that women genuinely aren’t very interested in science and therefore aren’t clicking on it, so it doesn’t show up… or is it that they are interested, but aren’t being shown it as an option in the first place?

She also suggests Stumbleupon could do with adding a “shuffle” button. Perhaps it would be suitable to have something like a “dismiss, I’m not interested” button, which would then replace the item with something else,. Similar things already exist in other places, like on the “people you may know” feature on LinkedIn.

Note: I haven’t had a proper look at the Stumbleupon site yet, this is based on a couple of tweets and Jo blog’s post.

Note: an update from @adam_j666 shows that the categories presented for men are slightly different each time the new user suggestion page is generated – for instance, “photoshop”, which was only offered to women in Jo’s case, now turns up for men, which would put it in the “both” category. A more rigourous analysis would involve requesting the page several times (the more the better), then counting how often categories show up as either “male” or “female”. I’ll get cracking.