My 1944 Longines History

Posts
2,587
Likes
5,644
There are always lots of duplicates with these records, usually on average for the ones I have used for myself, there have been averages of about 5 or so duplicates. The point is that the name is so common in the USA, that finding one purely on the basis of name without an address or some uniquely identifying fact, even in the same city you might as well be looking for John Smith. I have to trust that the watchmaker met someone with the name of Ed. Why he would give a personalized gift from his parents to a watchmaker, it is a mystery.
It’s always hard to know what sources to trust, but Corey is a pretty unusual surname and there seem to be only around 20,000 currently in the US. If you strip out the ladies, that’ll drop to 10,000. The Edwards are only going to account for few hundred, so by the time you add the middle initial I guess you’d be down into low double digits.
 
Posts
1,446
Likes
4,405
It’s always hard to know what sources to trust, but Corey is a pretty unusual surname and there seem to be only around 20,000 currently in the US. If you strip out the ladies, that’ll drop to 10,000. The Edwards are only going to account for few hundred, so by the time you add the middle initial I guess you’d be down into low double digits.

I don't think that is correct. Could you show us that? My search was specifically on "Edward M Corey". Did you see the search results was 628,582 records for only the United States.
x1428142-b13ef4a884abe29d797218552c867c74.jpg.pagespeed.ic.X3y4XjeLFb.webp
 
Posts
2,587
Likes
5,644
I don't think that is correct. Could you show us that?
Which part ?
 
Posts
2,587
Likes
5,644
That may be so, but just think about the number. It make absolutely no sense.

https://www.ssa.gov/oact/babynames/decades/names1920s.html

Here we can see that a boy born in the 1920s would have about a 1:50 chance of being an Edward. Assuming middle names are distributed in a similar manner to first names, a quick estimate give also around a 1:50 chance of having the middle initial M. So 1:2500 approx 20 year old males would be Edward Ms. Given the US population (at around 400 per million) it would be around 50,000 in total. That’s for all surnames !
 
Posts
2,587
Likes
5,644
What happens if you search census data rather than obituaries ?
 
Posts
1,446
Likes
4,405
There are a lot more people that immigrated to the United States, how do you account for those people. What information do you have about the additional population? It is not a closed system and where are all the assumptions coming from? The only thing that we know is how many public records exist.
 
Posts
1,446
Likes
4,405
I don't have an Ancestry.com subscription. Would you share with us any results if you do do the search?
 
Posts
2,587
Likes
5,644
As my guess, and it is a guess, is about 4 orders of magnitude away from yours, I don’t think the reason is unrecorded immigration.
 
Posts
1,446
Likes
4,405
Looks like my free trial is over and I don't see a way to get to that search screen without a subscription
 
Posts
2,587
Likes
5,644
I don't have an Ancestry.com subscription. Would you share with us any results if you do do the search?
Neither do I, but I seem to have got a free search… and this is not including the middle initial.

 
Posts
1,446
Likes
4,405
Even if there were only 5 Edward M Corey in the United States. You would need a lot more information to pick out the exact one.
 
Posts
1,446
Likes
4,405
I think you need to look at it from 1920-1950. The census wasn't really getting around in the 1920s. You need to see during the 1950s when the records were much better and then use the date of birth. Otherwise you are basing the population from very early Census Bureau data that did not cover the whole United States.
 
Posts
2,587
Likes
5,644
Even if there were only 5 Edward M Corey in the United States. You would need a lot more information to pick out the exact one.
This is true. I was only trying to understand the reasoning for the quoted number of 600,000. I think using obituaries as a counting method is very questionable, so it’s good that others know the risk. Census data seems much better, and gives values that are intuitively in the right ballpark.
 
Posts
1,446
Likes
4,405
1920s, many of the US government organizations did not exist yet.
 
Posts
2,587
Likes
5,644
I think you need to look at it from 1920-1950. The census wasn't really getting around in the 1920s. You need to see during the 1950s when the records were much better and then use the date of birth. Otherwise you are basing the population from very early Census Bureau data that did not cover the whole United States.
Yes but I’m generally scaling up from percentages to the total population. I guess that removes most of the risk.
 
Posts
1,446
Likes
4,405
1920s many of the records were Church records or Township records which unlike Europe, not one really cared.
 
Posts
1,446
Likes
4,405
Yes but I’m generally scaling up from percentages to the total population. I guess that removes most of the risk.

I'd like to strongly disagree with this. The records were not kept by any central organization at that time and finding information for that era really depends on finding local newpapers.
 
Posts
1,446
Likes
4,405
By 1950s many of the government organizations did have good records because of WW2 and the draft. The draft was how eligible males were required to go to war.