Some Genealogy Discrepancy Research Tips


By Chris Mills

I'm going to focus this article rather narrowly on trying to work out discrepancies between dates and names you find on burial markers (something I'm really familiar with from all the work I've done on Find A Grave) that don't match other records you find for the same person, or who you at least THINK is the same person.

Which Year Is Correct?

Contrary to popular opinion, grave markers are not nearly as infallible as people would like to think they are. And when you think about it, it makes perfect sense that they would be less than perfect sources of information.

Markers are sometimes ordered in a hurry, and the information is sometimes provided by someone who isn't even a family member. They may know the age of the person but not their exact birth date. Usually the death date or year is accurate, but not always. I have seen a lot of "rounding" type errors on markers where someone who was born late in a calendar year had their birth year listed as the following year. I have even seen cases where someone died close to the end of the year, and their death year was erroneously listed as being a year ahead, although sometimes it was the year before. My recent favorite was a memorial I found where the marker showed both an incorrect birth year and death year.

How Do You Even Know There Is A Discrepancy?

You know there is a discrepancy when you have access to other records that contradict what is on the marker. These records can be state death indexes, the social security death index, cemetery records, obituaries, birth indexes, family records, etc.

Which Date Is Correct?

Here's the problem. If you only have two sources, you have no way of knowing which of the two is correct. The only exception to this is the social security death index, because of the way the records were generated (access to originals or certified copies of original birth records or other similar records which prove date of birth). If someone is not listed in the social security death index, your best bet is to find a plurality of matching sources (i.e., find two or more sources that agree on a date). There are not a lot of entries in the social security death index prior to the 1960s, and I still find people that died in the 1990s and 2000s that do not show up either, although most people that have lived in the U.S. recently do show up.

Census records are not good for exact dates (although the 1900 census listed birth years and months, in a disturbing number of cases the birth years are off by one year, although some are off by more than that). However, the census records can help you in certain situations. If someone was supposed to have been born in 1879 according to one set of records and 1881 in another set of records, if you can find them listed with their family in the 1880 census, you know the 1881 birth year is wrong. Conversely, if they were supposed to have been born in 1879 and you CANNOT find them in the 1880 census, it doesn't prove the later birth date, but it does at least cast a reasonable doubt on the earlier birth year.

How Do You Know You Have the Right Person At All?

This does bring up the disturbing question, if you have a person where you just can't match the information exactly, how do you know you have the same person?

In some cases, depending on how sloppy someone's research is, they may make a mistake and think that someone who is buried somewhere is a different person. Even people who know what they're doing can make some assumptions that prove wrong. For example, Find A Grave has placed a famous memorial in Forest Lawn Glendale (California) for the actor J. Lewis Smith, who died in 1964, complete with a marker photo that looks plausible. There's only one problem -- there is no evidence that the actor is actually buried at Forest Lawn, and the marker photo they posted is for a non-famous person named Lewis James Smith, who also died in 1964, four or five days after the actor. My experience with getting Famous Memorials on Find A Grave fixed has not been confidence-inspiring, so I have left this one alone. But my point is that you can know what you're doing and still make a mistake and get two different people mixed up, and think that someone is buried at a cemetery whom isn't there.

Sometimes this is a tough nut to crack and if the person has family members you know the names of you can cheat a bit by trying to find out if anyone buried near them is family you would expect to find, such as parents, spouses, children, siblings, etc. That's one way that Find A Grave is really useful, you can put out a photo request and ask the person who is attempting to get the marker photo to look for people with the same surname or specific first names who have markers near the person you have put the photo request in for.

Now, all of these different tactics I am talking about require some time to do research. That's fine if you are looking for a relative or someone you have some particular interest in. What if you took a marker photo just because you were walking through the cemetery and happened to take some pictures? You go to add the marker photo to Find A Grave and find an existing memorial but when you go to add the photo you notice some discrepancies. You're really not interested in doing the research. Unfortunately, in a lot of cases neither is the person you send the note to about the discrepancy. In a case like that if neither of you wants to research it the best thing that can be done is for the memorial owner to put a note in the bio section of the memorial that there is a discrepancy between the dates in the paper or online records and the marker.

In some cases if you are unsure about someone you can call or email the cemetery you think they are buried in and try and get more information. Unfortunately, some cemeteries are now charging money just to answer questions to help you determine if a relative is buried in their cemetery. I think this is an appalling practice and I refuse to research any cemetery that does this, unless I can actually go there and take marker photos without the help of the staff (and at that, those cemeteries are more likely to try and run you off of the grounds if they catch you taking more than a couple of photos).

By the way, if you think these points are not worthy of discussion (i.e., who cares about a difference of a year or two on a marker?) it's not the one or two year differences that get me bent out of shape. I have seen discrepancies of as much as 10 or 12 years between death index records and dates on markers, and those scare me, those are the kind of things where you really wonder if you have the correct person or not. In the cases I have researched, they HAVE been the same person, but it does make you wonder. By the way, one of my grandfathers has a marker that lists his birth year as being five years earlier than what it actually was. He lied about his age when he joined the Army and apparently because he got a military marker provided by the government on his grave, they used his "official" birth date that he gave them when he joined the Army. I know it's wrong. You know it's wrong now too. It's still on the marker wrong, and I'm guessing it's not going to get changed anytime soon.

Occam's Razor AKA Common Sense

I'm just going to throw out a few more things here. If you've read this far, you're obviously a glutton for punishment and can handle a bit more verbiage. Occam's Razor is a proposition put forth by William of Ockham, a medieval English philosopher. Simply stated, it is thus: If all other things are equal in a question of logic that involves two or more different explanations for a phenomenon, the simplest explanation (or simpler one) is usually the correct one.

What that means in this context is that if you are looking at a memorial and it lists a birth date of December 26, 1899 and the marker shows a birth year of 1900, the actual date you are seeing is likely the correct one, and the year on the marker is an approximation. Likewise if you see a birth or death date of January 3, 1915 and the marker shows a year of 1914, ditto. There are two exceptions to this which I will note below.

Sometimes the birth records you are looking at are based on a christening date rather than an actual birth date. Likewise, sometimes the death records you are looking at show a burial date instead of an actual death date. It is incumbent on you to know which type of records you are looking at. And if you are looking at information someone else posted, hopefully they have documented if the dates they are using are based on christening or burial dates rather than actual birth or death dates.

The California Death Index and the years 1900/1901

I have never found this documented properly anywhere but I have found a few people complaining about it on genealogy message boards, so I know I am not the only one who has been brought to grief by this. In my particular case it caused me to duplicate thousands of memorials on Find A Grave and has also caused me hundreds of hours of work changing memorials that were already created.

One of the pieces of information recorded in the California Death Index (1940-1997) is the birth date of the deceased. There is a bug in the recording which, as far as I can tell, may have corrupted the birth year for literally hundreds of thousands of death records. I have no idea what caused it to happen, I have to assume at some point they had a problem recording dates that ended in two zeroes (i.e., 1900). I know this was a systemic bug because I have gotten several hundred correction requests on memorials I created with a birth year of 1901 where the marker showed 1900, and in my own cemetery walking I have discovered dozens (just from my own more limited sampling) of memorials I created with the same birth year where the marker showed 1900 instead of the 1901 that I had recorded.

At some point after I realized there was a problem I did some searching on the California Death Index with some pretty common surnames to have a fairly large list of results to do some crude analysis on. These were surnames like Smith and Miller, which seemed to work fine for this purpose. My test was rather simple. I searched for that surname and then put in birth years in a range from 1895 to 1905 and then counted the number of matches I got on each search. I recorded the results somewhere but don't have the numbers handy, and there is no need to bore you with that level of detail.

The results confirmed by suspicions. The years 1895-1899 and 1902-1905 all contained roughly the same number of matches on the common surnames I queried. However, there was a huge dip in 1900 followed by a corresponding spike in 1901. There were only 10% as many names listed in 1900 as there were in 1901.

Again, I have never found this documented anywhere, but I think the numbers speak for themselves. There was a huge bug in the California Death Index for those years. The relevance it has for me is that since I was using the California Death Index as a source for hundreds of thousands of memorials I created on Find A Grave, if you see any memorial I created where my memorial shows a birth year of 1901 but you see a marker photo that shows 1900, you can pretty safely assume the birth year was actually 1900 like the marker photo shows.

I was going to go on for a bit with some musings on names (first names, middle names, nicknames, surnames, prior married names) and some of the shortcomings of the way Find A Grave does or doesn't allow you to search on these fields, but I am realizing this particular article has probably reached its logical limit, both in terms of my time spent writing it and the time of anyone else who might be reading it. If I write another article on name discrepancies I will append a link to it later at the bottom of this article, and in the meantime leave you in peace for your other pursuits.

Page created 14 September 2013

Online resources (free unless otherwise noted):

California Death Index 1940-1997 on Family Tree Legends
California Death Index 1940-1997 on Rootsweb
Social Security Death Index on Family Tree Legends (only has entries through part of February 2003, nothing later than that)
Ancestry.com Social Security Death Index (subscription required)
Other California Death Index Links on Chris's website (includes death indexes prior to 1940)
Regional Genealogy Links on Chris's website (includes links to other state death indexes such as Illinois and Missouri

Return to Genealogy Links
Return to Main Page

Contact us

© 2013 by Chris Mills