Fun with small data sets. As far as I can tell, fully 1/3 of the difference they describe is due to a single female judge with the unlikely name of "Bruce". This line is my favorite:
> Moreover, changing a girl’s name from “Sue” to “Bruce” (99% of those named “Bruce” in South Carolina’s voting population are males) increases her probability of becoming a judge by a factor of 5 (roughly).
Right. And changing a boy's middle name to Hussein no doubt increases his changes of reaching the White House by a factor of 1000.
"Masculinity" is defined as "number of men who have this name per number of people who have this name." Seems to me like A)it's difficult to do this test accuracy given the small number of people who have a given name and B)it doesn't necessarily accurately reflect how we perceive the masculinity of a name.
Also, "Because the number of female judges in the data is small, the asymptotic justification for conducting standard hypothesis tests is suspect."
This is an interesting paper, but it seems like it would benefit from a larger dataset.
"8. At the request of an anonymous referee, we perform a check on our nominal
masculinity measure by examining the score for Bacon Magazine’s “Top 10 Stripper
Names.” In theory, female exotic dancers choose hyper-feminized stage names. Only two
of those names, Candy and Porsche, had a nominal masculinity of 0. Three other names
had nominal masculinity names below the mean female voter. Two other names on the
list actually scored quite high in nominal masculinity; Angel had a nominal masculinity
of 0.15 (due to its popularity among Spanish speakers as a boy’s name) and Houston had
a nominal masculinity of 0.98. These findings suggest the potential for further research,
which is beyond the scope of this paper."
13 comments
[ 2.6 ms ] story [ 45.1 ms ] thread> Moreover, changing a girl’s name from “Sue” to “Bruce” (99% of those named “Bruce” in South Carolina’s voting population are males) increases her probability of becoming a judge by a factor of 5 (roughly).
Right. And changing a boy's middle name to Hussein no doubt increases his changes of reaching the White House by a factor of 1000.
On the other hand, it also vastly increases his chances of getting captured and hanged due to actions of the White House.
Also, "Because the number of female judges in the data is small, the asymptotic justification for conducting standard hypothesis tests is suspect."
This is an interesting paper, but it seems like it would benefit from a larger dataset.
"8. At the request of an anonymous referee, we perform a check on our nominal masculinity measure by examining the score for Bacon Magazine’s “Top 10 Stripper Names.” In theory, female exotic dancers choose hyper-feminized stage names. Only two of those names, Candy and Porsche, had a nominal masculinity of 0. Three other names had nominal masculinity names below the mean female voter. Two other names on the list actually scored quite high in nominal masculinity; Angel had a nominal masculinity of 0.15 (due to its popularity among Spanish speakers as a boy’s name) and Houston had a nominal masculinity of 0.98. These findings suggest the potential for further research, which is beyond the scope of this paper."