In line with a previous post I decided to have some more fun with names.
I used the census names data to generate 200 names by taking a random first name and a random last name and combining them. The first names were chosen 50/50 from the male and female lists. All the random choices were done uniformly so that I’d get “interesting” names instead of a bunch of Dave Joneses.
For each of the names I asked five mechanical turkers to judge the person to whom the name belongs on six axes:
- Hero or villain?
- Trustworthy or untrustworthy?
- Attractive or unattractive?
- Sophisticated or naive?
- Powerful or weak?
- Outgoing or reserved?
There were a few methodological issues that I need to sort out at some point. And I’m going to keep the full results secret for a little longer but I thought I’d share a few quick aggregate results:
- Most heroic – Kurt Rollin
- Most Villainous – Toney Prus
- Most outgoing – Jeanie Bainard
- Most reserved – Jesusita Fondaw
- Most trustworthy – Laureen Hodgen
- Most untrustworthy – Robbie Holck
- Most sophisticated – Cecille Legat
- Most naive – Ramon Smialowski
- Most powerful – Rich Keaty
- Most weak – Bonnie Wurz
- Most attractive – Whitney Okano
- Least attractive – Freddy Doughtery
How do you guys and gals read these names (and other ones)? Let me know in the comments!
Please Scoop Me! 
I wonder how useful Mechanical Turk is for these culturally determined opinions given that it’s difficult to control which country they’re based out of. For some reason, I get a vision of Chinese WOW farmers.
That’s a valid concern. But there have been lots of surveys of turkers to address this question (e.g., here and this one by my labmate.) and they have largely allayed my concerns.
It would be interesting to try the same test on a smaller set of names with a larger test group, in order to get statistical significance on which first/last names fall on the extremes of the different axes. I would also be curious how these results vary over time, and how they correlate to the most popular baby names of the time.
You’re right that with such small sample sizes it’s hard to gauge statistical significance. As I mentioned earlier there are some methodological issues that need to be ironed out first so take the post with a grain of salt.
Correlation with time is definitely an interesting aspect. I wonder what the perspective on a name popular in the 50s is relative to a name popular in the 90s. Alternatively, I could ask turkers to guess the age of a person based solely on their name.
From my backseat driver’s perspective, I think a lot of problems are going to come from the fact that first and last names are not independent random variables. Spammers use the same “census data, random first name, random last name” method with no accounting for name origin. So when most people see grossly implausible ethnic mismatches, the number one association it rings in their heads is “Viagra”. E.g. “Ayotunde Polkowski”, caught by my spam filter today.
Also, the main signal in judgments of “most attractive” and “age” may come from the co-occurrence of two names, rather than the names themselves. To give a concrete example, a lot of Asian immigrants and their kids carry presidential or Biblical names that would sound like they belong to an 80-year old if you heard them with a European surname — so for anyone aware of these trends, “Ruth Chang” sounds far more attractive than “Esther Bainard” not because “Ruth” is a more attractive name than “Esther” but because they figure Ms. Chang lives in a college dorm and Ms. Bainard in a nursing home.
I wonder how people come up with these feelings. Maybe first/lastname being shared with a well-known public figure? (phonetic similarity?) Ethnic overtones?
Ethnicity is almost certainly a huge factor. It would probably not be too hard to do an ethnicity study as well and see what kind of correlations there are.
Just want to say your article is striking. The clarity in your post is simply striking and i can assume you are an expert on this field. Well with your permission allow me to grab your rss feed to keep up to date with succeeding post. Thanks a million and please keep up the effective work