Tuesday's Tip - A Tagging Tip (Intermediate)
Register for Webinar Friday - Full Circle: Tracing the Descendants of a Slaveholding Ancestor by Nicka Smith

Don't Rely on Indexes


I was searching for George Patton in the 1820 census and I found him in an index which made me happy. Here is an image of what was listed in the index as George Patton.

1820 Census Lydia Patton entry
(click image to enlarge)

On this very same page there was another person by the name of George.  You can clearly see what the name George should look like as written by this enumerator. The first image, showing the man indexed as "George" Patton, clearly isn't the name George based on the second entry.

1820 Census George Tilley entry
(click image to enlarge)

1820 U.S. census, Wilkes County, Georgia, p. 162, lines 4, 20, Lydia Patton and Georgia Tilley; digital image, Ancestry.com (https://www.ancestry.com : accessed 27 December 2017); citing NARA microfilm publication M33, roll 9.

Luckily I know who the first entry is. It is George’s stepmother Lydia (Orr) Patton. So here is the problem. Lydia and George are very different words so no matter how fuzzy you make this search these two will not be picked up as a possibility for the other unless I searched for the name Patton only. Let’s say I did search just for the name Patton. If you were looking for a Lydia Patton would you click on the name George in the index? Or, if you were looking for someone named George would you click on Lydia? This is just an example. Lydia wasn’t in the index at all since she was indexed as George.
Don’t get me wrong. I know that the handwriting is hard to read and it is always easier for someone who is familiar with the names to spot them. My point is, don’t rely on indexes. The indexers are human and they make mistakes. Sometimes you need to hand search the images.
All of the online repositories have this indexed incorrectly and I have sent corrections to all of them. This is how you can make the index better. If you see something like this let them know.
So where is George? No clue. I haven't found him yet.

Michele Simmons Lewis, CG® is part of the Legacy Family Tree team at MyHeritage. She handles the enhancement suggestions that come in from our users as well as writing for Legacy News. You can usually find her hanging out on the Legacy User Group Facebook page answering questions and posting tips.


Feed You can follow this conversation by subscribing to the comment feed for this post.

I had a similar problem with my James Beard, wife Harriet and a previously un documented/sourced daughter Elizabeth.
His was in the 1851 Wales census.
Finally found them under Bend. When viewed it was Beard and fairly clearly too. Still if I had not gone through page after page of images I would never have found them. Fuzzy doesn’t equate Beard and Bend.

Thanks for this always pertinent post.
I just wrote about a similar situation on my own blog. An indexed surname that was so far off, it would not be picked up in fuzzy searches. Instead of Thorp, the correct name they indexed Thersse.

I have found index entries for my great-grandmother's baptism in two independent places...still, I know better, so wrote to the repository that holds the record. Someone there confirmed it for me, but until I can get hold of a copy (complicated by the repository on another continent only accepting payment if I phone with my credit card info), I'm not entering this info as a certainty.

I found the Samuel Graden family with sons John, George and Wilhelm in the Pennsylvania 1850 census but I was looking for Samuel with sons John, George and seven or eight more children. Looking more closely, Wilhelm was a female the correct age to be Mahala, the daughter I was seeking.

By following the household numbers, I found that the census pages/images were out of order and the rest of the children were at the top of a page three images further along. Who knows how those "orphans" were indexed.

Excellent! I always tell people to look at several pages before and several pages after when looking at a census. Usually I say it so that people can get a feel for the handwriting or to find extended family members but you found that the pages were out of order (and I have seen this) which is something that most researchers aren't expecting.

Now I don't feel so bad. When I first started researching, I spent a few years trying to track down the homestead application for Philip Tkachuk or Tkaczyk, plus some other variants, then gave up on him and started looking for his brother Harry, who also had a homestead. Couldn't find him at first, then dropped the first name to do a general search in the same region using various spellings. Finally found a Hawrylo Tkaczuk that fit the information I had, being that Harry was his nickname. Began searching again using homestead locations, and lo and behold found a Philip "Skaczuk" close by, who was the correct one! It was a transcription error due to the scrawled signature. Never even crossed my mind to try jumping ship on the surname initial. Lesson learned.

Unusual names are often indexed as something more familiar. My Joah Carter is sometimes listed as John or Joseph (even spell check has Joah underlined as wrong). Lourinda (also underlines by spell check) gets listed as Louise or maybe Lucinda or another form. The suggestion of looking at more than just your ancestors census entry is great for other reasons - who were their neighbors? Are they related (a good possibility)? I have found old maps that show who lived where and have been able to see who else was part of their life, or how far they walked to visit friends and relatives or to get to school, church or the blacksmith and grocer. For one ancestor I have his diary, and he changes the name of his hired hand from Siras and several alternatives to finally settle on Cyrus, which makes more sense and turned out to be a relative.

Excellent detective work, Kathy!

I had an ancestor whose first name was Emmarilla. On every single census, death certificate or child's marriage registration on which she is mentioned. her name is spelled differently. Emerelda, Emmarilda, Emmarill, Emerilla, etc., thus the spellings in listings linked to her in online geneaology sites are all over the map. Her gravestone says Emmarilla, and I'm sure her husband would have made sure it was correct. Is there any way to correct these errors?

Here's another tip regarding indexes. If a data base, like the census, is available in more than one place, and you can't find your information in one, check another. Indexes are not always the same. I could not find a family in the 1930 census, but when I checked the same census in another source, there they were!

Recently I found 2 children in an Ohio Birth Index that appeared in no other record types but had the same birth date as 2 other known children. In checking the same Birth Index in another source I found that the 2 mystery children did not in fact exist. The State file numbers in the first index belonged to completely different people and those people in turn had incorrect numbers in the first index.

Even though the indexes may be correct, sometimes without seeing the original documents you can still make the wrong assumption.
These two entries in the BDM indexes show two people with sequentional registration numbers, leading a lot of people to assume they are twins.

Name Registration NumberFather's Given Name(s)Mother's Given Name(s)

Iza was actually born in 1887, and Roland was born 1890. the reason for the confusing dates was the law was changed and these were late registration of their births.

So, Michele Simmons Lewis, did you follow up and send a "please look at this again and consider correcting the index" based on what you explained here?

Yes. When I find errors in the index I send in corrections.

I found my ancestor in the North Carolina Tax index on Ancestry. Everyone on Ancestry trees has him living there in the 1780s. As I looked at it, I realized that the county and district names were identical to where I had found him in Georgia in the early 1800s. I went page by page in the Georgia records--and there he was on the date and page the index said was in North Carolina!

You really need to see the originals in any index.

This is why I get very upset with Ancestry, Family Search, Find My Past and others who only have typed out information. Where did they get the information and why not gives us the original and now in most subscriptions we pay for all information and it could be incorrect.

They have to have permission from whoever holds the originals to be able to put the images online.

What infuriates me about Ancestry is that they absolutely refuse to change any of their mistakes. I constantly send them corrections due to Transcriber error, but they just note it as an alternate. It makes me wonder how many people did not get the correct information.

I have had Ancestry change some erroneous index entries. Maybe it depends on who handles the case?

The comments to this entry are closed.