Sunday, July 12, 2009

Uncommon names correlated with increased juvenile delinquincy

A study by David Kalist (also registered at IDEAS) and Daniel Lee of Shippensburg University seems to indicate that unusual or uncommon names are correlated with increases in juvenile delinquency. ("First Names and Crime: Does Unpopularity Spell Trouble?" - full text in html, abstract).
The study takes the list of names from an undisclosed state and creates a Popular name index which is essentially the ratio of the popularity of a name to the most popular name in his list, Michael.
A cumulative distribution plot of the popular name index (PNI) for the two populations, the state population and the juvenile delinquents shows that the juvenile delinquent names are less popular than the general state population.

The plot shows a considerable difference in the medians (50th percentile). About half the names in the state population have a PNI greater than 20, while about half the names of juvenile delinquents have a PNI greater than 11. The average PNI for the state population is 26.31 but falls to 22.0 for the juvenile delinquents. The study says that this difference is significant. Therefore, compared to the state population of names, a larger proportion of juvenile delinquents have unpopular names.

The study also correlates the PNI to the fraction of the population that is a juvenile delinquent. It shows that a 10 percent decrease in the popularity-name index (less popular name) increases the number of juvenile delinquents by 3.67 percent.

Of course, the comments at Neatorama and even by supposedly educated researchers in the newspaper article spout off about examples of people they know that were against type according to the study. One should remember that the plural of anecdotes is not data, it is anecdotes. The study is over a population with a distribution, you will always be able to find examples of youths with unusual names who are not delinquents and those with common names who are, you would almost have to, given to have the results be balanced. That doesn't necessarily invalidate the study. I guess people like talking about that one friend from childhood with the certain name that was trouble, or even that one friend who was trouble no matter the name.

That being said, correlation is not causation and so the results of the study should be examined critically as with all scientific study. There is some evidence that the choice of names is correlated with other socio-economic and racial indicators which themselves are sometimes correlated with increased crime or juvenile delinquency, but the question remains, does the name make the person or do are they both caused by some other factor.

(via Neatorama, via Orlando Sentinel, via Social Science Quarterly)

No comments: