Wednesday, July 14, 2010

Reporting Indexing Errors

Record view - links for report or fixing errorsDear Ancestry Insider,

I am in the FamilySearch indexing project, and am very worried about the incidence of errors in indexing. PLEASE, tell me how I can report the great number of indexing errors I find while using my account to search U.S. censuses?

I see the link to "report problems", which refers only wrong images or unreadable images - but I need to report obvious errors in indexing as compared to the information easily read on images!

Thanks, M. A. Farrell


Dear M. A.,

You are correct. “Report Problem” (in the image viewer, below) and “Report Image Problem” (in the record view, above) are used to report missing, wrong, or unreadable images.

To report errors in indexing use “Add Update” (below) or “Add Alternate Information” (above). Enter your interpretation of what is written. Yes, expects you to provide your idea of the correct information. Handwriting interpretation can be a matter of opinion. While it may be obvious to you, keep in mind that it has already been misread by once. And just in case your opinion is wrong, will keep both the old and the new values.

I hope that helps.

-- The Ancestry Insider

Image view - links for reporting or fixing errors


  1. I probably add 15-20 corrections a week to Ancestry ... this helps others, but also helps me ... I receive quite a few inquiries from people who have read my corrections. I use the term 'corrections' liberally and often add date/place of birth or a missing middle name ... even if there is nothing incorrect about what is shown.

  2. I have made corrections on census records on Ancestry. And they were posted. Since I know the family it was obvious to me what the real name was. But in looking at the record I can see how it wouldn't be as clear to someone who was indexing.

  3. Adding corrections also turns out to be a great way to get in touch with others researching the same family, so this is additional incentive to submit those corrections.

  4. Unfortunately there are many sorts of errors that cannot be reported except as 'comments', which are not indexed.

    The alternative readings that are allowed (at least names and ages in the case of USCensus reports) do eventually show up in the index.

    But where the extract (so-called "record") lists people in a household who are not even on the same or adjacent page there is no immediate way to differentiate them.

    The more serious issue is where has invented relationshps between Census household members that are not given in the enumeration at all. Since these mistakes are encoded, it is all to easy to corrupt tree data. Too many clickophiles don't bother to read the actual document, or may not have a working knowledge of what actually was enumerated.

    For example it is common for to code the wife of head of household as mother of a hh's child (so aggravating in the case of 2nd, 3rd, 4th spouse), and for the extractor to have selected a household member to be parent of a hh's grandchild.

  5. You say, "To report errors in indexing use “Add Update” (below) or “Add Alternate Information."

    No such critter for census images.

    For the index of 1840 Census, Vermont, Bennington County, Arlington:

    The indexers have the township listed as "Helington" instead of Arlington.

  6. You wrote "To report errors in indexing use “Add Update” (below) or “Add Alternate Information."

    No such critter for census images.

    The indexing for 1840 Vermont, Bennington County, Arlington has:

    "Helmington" instead of Arlington.

  7. Equally (if not more) aggravating are the errors in place names. I have relatives who migrated to what became known as the town of Mequon, Washington Co. (later Ozaukee Co.), Wisconsin (Territory).

    Just try and find "Mequon" as a place name on Ancestry's 1840 and 1850 federal censuses! It's true, the original enumerator in 1840 wrote something like "Mignon" at the top of the census pages, and the 1850 enumerator might have as well; they were probably spelling phonetically. But the place has never been known as "Mignon"--not on any map or gazetteer, newspaper or federal or state document of the time. The enumerator and/or the transcribers got it wrong.

    So, if you are looking for hard-to-find relatives in "Mequon" you won't get their early census records unless you know to go to the county level (and you know that this part of modern Ozaukee county was once part of Washington county) and drill down to the town level and figure that, for Ancestry, Mequon is sometimes listed as Mignon.

    Sorry for the long example. I have many such in my researches. The main issue is that Ancestry provides no way to get these misleading place name listings corrected. You can't even list "alternative" place names.

    FWIW, I sent emails and blog posts to Ancestry about this specific problem over two years ago and--of course--I never heard back and the problems have not been corrected.

    Par for the course, alas…


  8. Even census-taker handwriting in the 1880 Orangeburg County, South Carolina has been a very serious problem for Ancestry as they have not been able to read surnames such as Hungerpiller that connect to my Perkins family down there. Hope that others who know my state like I know the Palmetto State will continue to advise Ancestry of all of their census errors that confuse the novice researcher.
    Dallas L. Phelps; Camden, S.C.


Note: Only a member of this blog may post a comment.