We are looking for the number of people currently living in New York
going by the name of Peter, Pedro, Pierre, Paeder, Petra, etc.

Obviously, an exact number is unknowable.  We're looking for a
combination of hard data and statistical estimation.

One interesting problem here is the number of people with 'Peter' in
their middle name who 'go by' Peter - and not their first name.

There are several possible sample sets to mine: motor vehicle
registrations, tax filings, school records, and the various phone
books.

Would it be possible to merge these information sets, sample
them intelligently, and thereby arrive at a better or more accurate
answer?

---

Request for Question Clarification by pafalafa-ga on 13 Feb 2005 19:56 PST

querious-ga,

Curious question you've asked here!  Interesting, but curious...

Although databases like voter registration and DMV records could
probably shed some light on the "Peter" question, these data sources
are not available to the general public.

In fact, I can only think of one commercially-available data source
that has an extensive dataset of New Yorkers, and allows searching by
first names.

I checked this source, and it shows the following for NYC:

Peter -- 9,116

Pedro -- 2,435

Pierre --  608


and so on.   


If you list all the names you consider Peter-like, I can certainly
generate NYC counts for the names.


There is also a list of first names ranked by frequency that was
compiled by the Census Bureau based on the 1990 census.  It shows, for
instance, that "Peter" is #43 in popularity for male names, and that
0.381% of men in the US are named Peter.

This is a profile of names for the US as a whole, and there is
certainly going to be some geographic variability.  Nevertheless, it
could be used to more or less ground-truth the counts of names from
the database I mentioned.

Does this seem like a reasonable approach?  If so, let me know all the
names that you would like a count for, and I'll generate the numbers,
and check their frequency against the Census information.

Thanks,

pafalafa-ga

---

Request for Question Clarification by pafalafa-ga on 16 Feb 2005 16:43 PST

querious-ga,

Just checking in.  Did my Peter-list spark any interest, or do you
prefer a different approach?

pafalafa-ga

I've done some more digging on this (oddly intriguing) question of
yours, and here's what I've come up with.

A commercial database, Reference USA, is available at many libraries,
and is considered one of the largest and most reliable databases of
people in the US, containing the names and address listings for 120
million US households.

By accessing the listings for NYC by first name, I was able to
determine the following numbers for New York City:



PETE* (PETE or PETER)..........9,116


PEDRO..........................2,435


PIERRE...........................608


PIETR* (PIETRO or PIETRI)........163


PETR.............................138


TOTAL of Peter-like names.....12,460



==========


According to the Census Bureau, Peter is the 43rd most popular male
name, and Peter's account for 0.381% of the male population:


http://www.ci.tucson.az.us/planning/names.pdf
MOST COMMON NAMES IN THE UNITED STATES


based on an analysis of the 1990 Census.



The Census also tells us that there are approximately 3.8 million men in NYC:


http://factfinder.census.gov/servlet/SAFFFacts?_event=ChangeGeoContext&geo_id=16000US3651000&_geoContext=&_street=&_county=new+york+city&_cityTown=new+york+city&_state=&_zip=&_lang=en&_sse=on&ActiveGeoDiv=&_useEV=



Based on this, one would expect there to be 0.381% of the roughly 3.8
million men in NYC = 14,456 Peters in New York City.



This is greater than the number I arrived at, but still in the same ballpark.  

There are many potential reasons for the discrepancy, the two most
likely (to my mind) are:


--the distribution of names in NYC is not quite the same as the
national distribution

--the Reference USA listings -- as massive as they appear to be -- may
still be undercounting the full population of New Yorkers.


Nevertheless, I think this is probably as close as one can get to an
accurate count.


I trust this information fully meets your needs for the best-available
count of Peters.  However, please don't rate this answer until you
have everything you need.  If you would like any additional
information, just post a Request for Clarification to let me know how
I can assist you further, and I'm at your service.

All the best,

pafalafa-ga