Geneaolgy Research - Probabilty of Someone having the same name

[berksted]

I never posted in this forum, so if this questions is not relevant, please let me know. I have a family member trying to trace the biological father of my adopted grandfather. From his adoption records she has been able to find a person who matches his name - first, middle and last as well as the same birthday. The last name is Smith so it's common but this was also Minnesota in 1907 (population not too large). I think we have found the right person, but what are the odds of someone having the same name and birth date in a given year.

Since some names are more common that others, I realize this is difficult problem to actually compute, but how does one think about this type of probability?

 

[CRGreathouse]

It's a hard problem because names aren't independent. A "Smith" is more likely to be a "Lawrence" than a "Maurice" (though they're of similar popularity in the overall population).

Toward a first approximation, since Smith is a popular name, I would check records (whatever you can find... SS death records if nothing else, though use a later date in that case) to see what percentage of Smiths in Minnesota have a matching first name and what percentage have a missing last name. Multiply those probabilities together then by the number of people in your database (hopefully not the same one you used to get the percentages) to get a rough first guess of how many matches you'd expect through chance alone. If it's small, there's a good chance that's the one.

 

[zli034]

Randomly pick 6 people from the street. The probability for at least 3 of them have same name or at least 3 of them have different name equals to 1.

 

[jhonpeterson]

This is common problem, I read some other threads & this is so interesting, if anyone who is know more about it so please reply, and I found such useful http://www.kinematik.com/" for researchers & developers.