Yesterday in the car, Abby and I happened upon a conversation about how other people will perceive our daughters’ ethnicity. They’re both half White / half Chinese, something that in Los Angeles is hardly rare, and we were split on how people would see them as they grow up (Zoe is 4 years old and mostly the focus of the discussion).
Rather than leave it as a hypothetical discussion, something we’d wait years to see how it played out mostly through what Zoe chooses to tell us, I decided to do something more scientific. I had Abby pick a picture of Zoe and we wrote up a Mechanical Turk job to have people answer what ethnicity they thought the person in the picture was.
Mechanical Turk, for those who haven’t heard of it, is an Amazon service that allows you to ask a set of humans to perform some work on their computer and report back on the results. You pay them a small reward (we paid them $0.05 each) and pick how many answers you’d like to get (we picked 100). You formulate the work, in our case in the form of a picture with a multiple choice question, and Amazon gives it out to people that make money performing the work.
To my delight, it only took about an hour to get 100 replies, making this science experiment quite fulfilling. The results were also really interesting.
And the results:
I gave people the ability to pick multiple, that’s why the numbers don’t add up to 100. Of the people that picked multiples, no one picked the correct one either.
Ethnicity identification is very much a regional effect. It’s not possible to know where the answers were from, but if we assume they were mostly American (given the time that I was using Mechanical Turk), this provides an interesting insight into how people will see Zoe and Kira in their lives.
It was also a fun mini-experiment.