Starting Bootcamp

[kato13]

In case someone wants the data here is 1990 and 2000 (either for this project or a future one)

http://www.census.gov/topics/populat...namefiles.html

http://www.census.gov/topics/populat..._surnames.html

The 1990 data has count breakdown and first and last names

the 2000 data only has last names but has a racial breakdown of the last names.

I might merge the 1990 and 2000 last name data and then see if I can find a racial breakdown of first names. The reason I quit work on this before was that names like "Suk Gonzales", "Mohammed Wong" and "Jamal Bjorkman" would occur much more frequently than in real life. I know they are possible, (Bernando O'Higgins is one of my favorite names in history,) but if I am spending my time on a tool I push myself to be perfect.

My existing code would probably work for a group of 10 (you can create a fun backstory for for an odd name combination), but the original purpose for my work was to create a database element for every member of my project (from 10k to 50k depending on which version) so I wanted something more accurate.

[kato13]

Some other numbers I put together for my project (this data is like 15 years old so I am open to suggestions). Note this is for the entire project including support. (There might be more dependents in dedicated facilities, but these numbers were to reflect the children at prime and maybe regional bases)

Sex
Male 68.6%
Female 31.17%
Dependent 0.23%

Education
Dependent Child N/A
HS Dropout 0.7%
HS Graduate/GED 5%
Some College 12%
Certificate 12.5%
BA/BS 44.8%
MA/MS 12%
JD/LLD 0.5%
MD/DDS 1.5%
MFA 0.5%
PhD 3%
MBA 0.5%

I explain my high number of doctors and college degrees in general is the project providing tons of scholarships and focusing some of their recruitment focus on those candidates.

Service History
Special Forces* 1.5%
Combat Vet* 3%
War Era Vet* 7%
Military Service 13.5%
Police 3%
Civilian 72%

* Is not included in military service total military service is 25%

Home state
Alabama 1.61%
Alaska 0.23%
Arizona 1.56%
Arkansas 0.94%
California 12.03%
Colorado 1.40%
Connecticut 1.25%
Deleware 0.27%
DC 0.22%
Florida 5.34%
Georgia 2.70%
Hawaii 0.45%
Idaho 0.43%
Illinois 4.50%
Indiana 2.20%
Iowa 1.08%
Kansas 0.98%
Kentucky 1.46%
Louisiana 1.65%
Maine 0.47%
Maryland 1.92%
Massachusetts 2.31%
Michigan 3.63%
Minnesota 1.75%
Mississippi 1.02%
Missouri 2.40%
Montana 0.33%
Nebraska 0.62%
Nevada 0.56%
New Hampshire 0.43%
New Jersey 3.02%
New Mexico 0.63%
New York 6.95%
North Carolina 2.71%
North Dakota 0.24%
Ohio 4.25%
Oklahoma 1.25%
Oregon 1.18%
Pennsylvania 4.61%
Rhode Island 0.38%
South Carolina 1.40%
South Dakota 0.28%
Tennessee 1.98%
Texas 7.03%
Utah 0.73%
Vermont 0.22%
Virginia 2.51%
Washington 2.04%
West Virginia 0.70%
Wisconsin 1.94%
Wyoming 0.18%

[ArielT]

Originally Posted by kato13

I might merge the 1990 and 2000 last name data and then see if I can find a racial breakdown of first names. The reason I quit work on this before was that names like "Suk Gonzales", "Mohammed Wong" and "Jamal Bjorkman" would occur much more frequently than in real life. I know they are possible, (Bernando O'Higgins is one of my favorite names in history,) but if I am spending my time on a tool I push myself to be perfect.

So the census doesn't offer first names by ethnicity, hrm... Is it possible to get a list of full names?

[kato13]

I have not seen a full breakdown by ethnicity at a national level.

It is possible I could get the information with a request to the census bureau, but if this was easy I expect someone would have gotten it and shared it. There might be some weird governmental hoops to jump through. One time I was looking for oil well production nationally, and when I got to Pennsylvania their Bureau of Land Management told me I had to come into the offices where they would burn me a CD (needless to say it was not worth the 700 mile trip).

The best I have found so far is recent listings (2008+) is top (100?) newborn baby names in New York state by ethnicity. It is something but naming patterns have seen quite a bit of change since the 60s (when I expect some project members might have been born no matter what version you use) so I am keeping that as an option if I don't find anything better.

[kato13]

Did a little more research. You can get full census data for the 1940 census as data is released 72 years after the census is taken. Unfortunately the data is presented in images not data files (There is also no Hispanic breakdown at all).

Found a research paper that had an interesting way building the first-name/ethnicity data. They looked at names on wikipedia and what ethnic categories they were assigned to. Interesting and if I can find the results I will see what I can do with it, but I have only been able to find the abstract so far.