Yeah. Common saying amongst my friends, mostly techies.
"The plural of 'anecdote' isn't 'data.'"
To counter that, though, I would say that your presents to your nieces, while a great idea and well-encouraged, palls in the face of parental and social pressure. Their friends are playing with dolls. Playing with legos by yourself can be fun, but not quite as fun as playing with their friends.
Tonight one niece was playing dolls by herself. The other was baking by herself. shrug Actually data is comprised of anecdotes and anecdotes are data, in my judgment. Life and actual experiences in the real world are data. Reality is data.
If your point is, well, this is just two little girls, you're right, that alone would not allow one to extrapolate to all little girls everywhere. However, I know for a fact that large companies and institutions have already done lots of experiments with kids (especially companies that have a vested interest in figuring out exactly what kids want, in order to maximize their profit) and my understanding is that the studies show that yes, girls mostly prefer to play with girlish things and boys with boyish things. This was the result of Science with a capital S. Now, do all boys hate dolls and all girls hate robots and dump trucks and guns? Of course not. But in the majority of cases that's how the cookie crumbled in the real world.
Also the whole "pink vs blue" thing you see in stores in the kid sections? Do you think that's some conspiracy or cultural artifact? From what I've heard, it's isn't. They've done experiments. Girls in general, world-wide, really do tend to prefer pink and boys prefer blue. In the general case, when dealing with large enough numbers, etc.
The actual colours are irrelevant, they were reversed a hundred years ago, it's the association that's important though.
Small children have a phase where they want things to be they way they "should" be. If it's a girl, she will like girly things because All Girls Like Girly Things because That's That Way It Is.
It takes until puberty when children start wanting individuality instead, sometimes with hilarious results in the opposite direction instead.
Studies show that it's incredibly hard to fight these tendencies, the onyl thing you can do is try to make tech more girly, by having female role models and to promote them.
/None/ of that is indicative of a genetic role. There is no indicator /anywhere/ in there that there is some genetic factor that makes pink stand out more vividly to girls than blue.
(edit: in fact, as anecdote - girls supposedly have more tastebuds, yet professional cooking is heavily male dominated. Though women cooks can put up with just as much if not more than men. Physical traits have no bearing on this particular cultural quirk.)
Conversely, all of that /can/ be explained as cultural artifacts, amplified by years of exposure, an encroaching western monoculture, and a feedback loop of expectations. When you stick "pink" and "blue" in front of a kid and ask them what seems more girly to them, you're already presenting a pre-existing baggage of cultural conditions, be it through subconscious impulses, or because the kid's been around long enough to know what his or her peers prefer.
Let's not lose sight that scientific experiments have to be interpreted, yeah? The process of science isn't nearly so clean that we can reliably say that any one experiment isn't tainted by subjective bias. Half of the fights over any one theory in any field at all is because of it.
That's interesting because the difference between dolls and robots is one of the skin/themes placed on humanoid shapes (albeit, robots can take on other shapes and still be robots). In the common case, they both have the same, painted on non-interactive expression. Robotic dolls and doll robots -- what exactly is the choice here.
With my son, we have a lot of animal themed toys that come out of the box without an associated gender. This hasn't kept us from adopting gender specific names for them though (Mr. Monkey, Sophie, Boss Hog, Leeroy Jenkins, Charlotte).
I actually remember reading, recently, that the color division is cultural. Point in fact is that around 1900 and earlier, if color had any meaning at all, it was reversed. Pink, being close to red, was manly. Blue, being close to violet, was womanly.
That the roles are globally reversed now could very well be indicative of a newish cultural shift that has nothing to do with nature, genes or suchlike.
While trying to find the article I stumbled across Straight dope (http://www.straightdope.com/columns/read/2831/was-pink-origi...), that claims that this idea might well be wrong. I does however, I think, speak out against the idea that pink-girly and blue-boyly is something known to be inherent.
Although I guess the thing I read earlier might well be wong, I'm still not that convinced that colors are that gender based by default.
"The plural of 'anecdote' isn't 'data.'"
To counter that, though, I would say that your presents to your nieces, while a great idea and well-encouraged, palls in the face of parental and social pressure. Their friends are playing with dolls. Playing with legos by yourself can be fun, but not quite as fun as playing with their friends.