25.6.10

Baby Name Statistics, Part 1

My other blog, JRI, deals with a very serious application of statistics: football. Because I love stats so much, I will occasionally do some goofy-ass project in my personal time, and this is one of those. It is an area that is much less serious than football: naming your children. I resolved to analyze the Social Security Administration's annually updated database of most popular baby names.

Registration for Social Security numbers became compulsory in 1937, which means that every individual born in the United States since then has had their name preserved on the rolls of SSA. From a statistical standpoint, the data available to me is not so much a sample as it is literally the entire population of data points.

I took the largest sample size (the top 1000 names for each year for each sex) compiled the samples into a spreadsheet, and calculated descriptive statistics for each name in the 2009 list--mean, mode, median, standard deviation. These were measures across the entire period 1937-2009. Not satisfied with that, I then obtained the complete set of unique names that appeared in the top 1000 over that same period and calculated the same descriptive statistics.

Finally, I devised a power statistic that is intended to measure popularity and consistency:



Mu and sigma are used in their standard sense to represent mean and standard deviation, alpha and beta are arbitrary weights (in this case 0.5 and 0.5), and rho is the percentage of the total lists that the name is present on.

I present the top ten names for boys and girls that are least likely to sound stupid as hell in 10 years, or alternatively, least likely to get your kid made fun of:

Girls

  1. Elizabeth

  2. Mary

  3. Katherine

  4. Rebecca

  5. Maria

  6. Sarah

  7. Anna

  8. Catherine

  9. Jacqueline

  10. Laura



Boys

  1. Michael

  2. David

  3. James

  4. John

  5. Joseph

  6. William

  7. Daniel

  8. Robert

  9. Thomas

  10. Anthony



It is interesting to note that 7 of the boy names are biblical; Anthony is Roman; and William (Germanic) and Robert (French transliteration of a Germanic name) date back to the Norman conquest of England.

Likewise, Elizabeth, Mary, Sarah, and Rebecca are biblical; Anna is the Latin form of Hannah, a biblical name; Maria is both the Latin form of Mary and the feminine form of Marius (a Roman name); Laura is the feminine form of the Latin laurus; Katherine/Catherine is the name of one of the first Christian saints; and Jacqueline is the feminine form of Jacques, which derives from Jacob...a biblical name.

I can conjecture that new names have some difficulty displacing the traditional biblical names, and that they also have trouble gaining a similar level of vast acceptance. It is also important to note that biblical names largely dominated the popularity rankings for the first several decades of the lists, with Michael, James, and Mary in particular sitting atop the list for years at a time. Many of the other names spent decades near the top of the list; James and Joseph spent one year between them out of the top 20.

No comments: