Although I didn't actually figure out the pattern until a couple of years ago, being in the physical presence of other human beings makes some of my orthogonal skills evaporate or sharply deteriorate. This is most notable with spatial organization skills (you would not believe how much stuff I can pack away into limited space... when no one is watching...) but applies to many other things, too.

I can accomplish a great many things, especially unfamiliar things, only when there is no help available. To the point where there was a period in my teens where I was new to cooking, and could only cook if my mother was not home, because if she was home, I would find the need to run to her 15 times asking where stuff was, how stuff worked, why the recipe said X, etc. Sometimes she became too fed up with me to answer the questions, or wouldn't know the answers, or would be too busy to engage with me at all, and this would render me completely helpless to do anything but turn all the heat sources off and flee. If she was not home, it never occurred to me to give up and wait for her return before proceeding with my task, and I'd look a little harder, scrutinze the appliance menus a little longer, apply my knowledge of how food interacts a little more diligently - and would produce perfectly fine food. I will also not learn routes around a neighborhood unless I walk them by myself - I can walk the same route with company a hundred times and then have no idea where the hell I am if I go alone the hundred and first.

My poorly-founded speculation: Girls may be more likely to try to solve problems with help from fellow humans. This help is unlikely to be available in a controlled study. (It's also frowned upon or unavailable in a number of academic or professional contexts.) Controlled studies are, however, likely to have humans present, blocking the bits that step up and say "okay, I'm doing this by myself now".