1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. If you're looking for the LostCousins site please click the logo in the top left corner - these forums are for existing LostCousins members only.
  3. This is the LostCousins Forum. If you were looking for the LostCousins website simply click the logo at the top left.
  4. Coronavirus Corner - a place to share your hopes, dreams, and frustrations.
  5. Only registered members can see all the forums - if you've received an invitation to join (it'll be on your My Summary page) please register NOW!

Incorrect entries?

Discussion in 'Comments on the latest newsletter' started by JennyMB, Feb 10, 2021.

  1. JennyMB

    JennyMB LostCousins Member

    I don't suppose I will ever know if one of MY entries is one of the incorrect ones that might fail to win a prize in the latest competition but I would be mighty dis-chuffed if I am. Three of my entries (2 from the same household) were marked as incorrect, but I have checked them all several times. I clicked the links to open the Census page at FMP and then carefully checked Surname, Forename, Middle Name or Initial where given and Age. Everything matched exactly. What more can one do?
  2. peter

    peter Administrator Staff Member

    Entries are never marked as incorrect since there is no way for the website to know if they are correct or not. A red ! indicates an entry that should be checked, because it is a near match for another entry, but you can confirm that your entry is correct. If you confirm that the entry is correct the ! will be removed (even if you're wrong to confirm the entry as correct).
  3. Bryman

    Bryman LostCousins Megastar

    A great feature to give members a second chance to obtain a match but could/should it be made even better? Correcting the entry (or not) and then confirming that it is correct does not always result in a match when perhaps it should.

    I always check any of my entries that are flagged with a red ! warning and I assume that other members do the same. Surely none of us dismiss such warnings without checking and assume that the other person has made an error. Once one member has confirmed that their entry is correct and the red ! has been removed, does it follow that the corresponding warning will immediately disappear (or at the time of the next search) for the other member as well?

    I have about 40 warnings every time that I perform a search and if the above is true then other members must have their corresponding entries similarly marked. When confirmation is indicated, either a match will result or the possibility will disappear with no record of a near match ever having been suggested. Presumably, one or both members are then likely to miss any subsequent match with a further member, or would the whole process repeat all over again from scratch?

    I am therefore reluctant to make any confirmation for fear of losing any indication of a near match having existed. I think it unlikely that near matches are completely spurious. Are other members taking similar (in)action for the same reason? Would it help if the red ! were not removed but simply replaced with a grey ! indication of previous warning? Although that may benefit me and others, I am a bit worried that it could affect processing elsewhere in the LC system and I am not wishing to add to Peter's workload.

    Alternatively, if the near match is to be discarded, should each involved member be informed of what the other member has entered (but not who the other member is!)? That could help where the source document is open to interpretation, as with poor hand writing on the original census form or use of different transcriptions.

    In the case where the 1881 E&W census is the source, we are told to use the Ancestry transcription rather than the one from FMP but the grey arrow always shows the FMP version. Is it possible to display the Ancestry version instead for the 1881 E&W census?
  4. Pauline

    Pauline LostCousins Superstar

    When I get a red ! I always check the entry and then confirm it as correct - having made any necessary changes, of course.

    It's easy enough to check entries in the 1881 transcriptions but as Bryman points out, the Ancestry version should be used for this and many people take the information from the FMP version.

    With some other censuses we take our information from the handwritten images, which not everyone has access to, and not infrequently the transcriptions are clearly wrong. I was adding an entry from the 1911 census yesterday where one forename was very clearly spelt Abrham and another Llily. These had been transcribed by FMP as Abraham and Hily, the former no doubt being what it was meant to be, but is not what was written on the form.

    Also, sometimes it is not clear what was the handwriting says and we can only make a best guess, and another person's best guess may be different, or they may be guided by a transcription. In such cases I hope that any match will at least be made via other members of the household.

    I'm not sure there is any easy answer to this, but at least we do get a warning where there is a possible error, and maybe most of the time these things are not a problem.
  5. peter

    peter Administrator Staff Member

    Confirming the entry as correct can never result in an immediate match. Changing an entry might result in a match but there is no way this can be guaranteed; if you have two near matches in the same household it is possible that they have been matched with each other.
    It might be possible, but it would be technically more difficult and involve more computation; there are also several advantages in using Findmypast for this census, one of which is that we use Findmypast for the other England & Wales censuses (so users are likely to be logged-in at Findmypast anyway). If an entry is flagged as a near match and your entry matches the Findmypast transcript, it is a good idea to check what Ancestry have in their transcript.

    Near matches usually involve differences in forenames and since the average household has 4 or 5 members the chance of a match being missed altogether is very small.
  6. peter

    peter Administrator Staff Member

    The census images are free for every census where we use them, other than the England & Wales censuses. For these censuses you need a subscription anyway, as the transcript isn't free - the free search results do not include the census references.

    This is, of course why members are strongly advised to focus on the 1881 Census - the chance of finding a match that has not been, or could not be, found through the 1881 Census is quite small.

    Interpreting handwriting can be very difficult for a transcriber who has no connection to the family, but it is much easier for a user whose family members they are.
  7. Pauline

    Pauline LostCousins Superstar

    I agree, but it does it happen sometimes and I have made matches on an 1911 or 1841 census entry that I haven't made by entering the nearest related people from 1881.
    Again I agree, but the errors I mentioned above - along with an 'Alfread' mistranscribed as 'Alfred' - were easy enough to spot without any need for family knowledge.
  8. peter

    peter Administrator Staff Member

    So have I - but it would have been made through the 1881 Census had my cousin followed the advice to focus on that census. Clearly anyone who can't access the 1841 and 1911 censuses will follow the advice, so problem solved.

    Note: I can't do anything about the fact that the 1841 and 1911 censuses aren't free online, but most members who don't have a subscription to any of the major sites are likely to have access through their local library, record office, FamilySearch centre, or family history society.
  9. Pauline

    Pauline LostCousins Superstar

    You can see free transcriptions for E&W censuses at FamilySearch, mostly (as far as I could see) with the full reference details included.
  10. peter

    peter Administrator Staff Member

    You're right - they have now added the census references for the 1841 and 1911 Censuses. But we can't switch from using the handwritten images to using the Findmypast transcription of those censuses because well over a million entries have been entered on the original basis.

    Of course, there's nothing stopping members using FamilySearch as their source and hoping that the transcript is sufficiently accurate - but they would still do better to focus their attention on the 1881 Census. Tracking relatives from 1841 through to 1881 and entering them and/or their descendants from the later census increases the chance of a match by around 10 times.
  11. Katie Bee

    Katie Bee LostCousins Member

    I cannot find anywhere on the Lost Cousins 'Read this First' or 'Help and Advice' that says use the Ancestry transcription.
    It all seems to say use the Findmypast website as the transcription is free.
    This is taken from Read this first: Step Two enter your relatives:-
    Note: the most accurate transcriptions of the 1841 and 1881 Censuses of England & Wales can be found at findmypast.com (and there you can search the 1881 Census free).

    In the Help and Advice section - getting started.
    The links to the 1881 census at Family Search and Ancestry do not work, but the Findmypast link did work.
    Maybe the links have changed and need updating.
  12. peter

    peter Administrator Staff Member

    I advise using Findmypast for the 1881 Census because only they provide a household transcript that clearly sets out ALL the information members need to enter - the very small chance that there is difference between the two sites is outweighed by other improvements in accuracy. I use Findmypast for my own 1881 households, so it would be disingenuous to recommend a different site to other members.

    However there is a Getting Started guide on the Help & Advice page for those who prefer to use Ancestry.[/QUOTE]
  13. peter

    peter Administrator Staff Member

    Thanks - the links in the guides do need updating - I tend to rely on members to point out when something is out-of-date as it's not feasible for me to keep track of every change. The same applies to links in old issues of my newsletter - I update them when I notice, but I don't go looking for them because there are hundreds of issues and thousands of links.
  14. Bryman

    Bryman LostCousins Megastar

    Are you sure? I corrected a forename recently and immediately checked via a repeated search. I found that the red ! had been replaced with a red tick for me but of course I know nothing about the other member, or even if they had even seen the red ! warning.

    I agree for most instances but households with a single member recorded would not have the fallback to other members for obtaining the match.
  15. lindy

    lindy LostCousins Member

    After reading the newsletter I decided to check my latest additions for errors. I use the 'grey arrows' when entering relatives so all the census refs were correct. But I disappointingly found I had made some errors in the names, and I thought I was so careful! Misspelling a forename and a family name, in other words putting in what I had expected to see in the transcript. And a couple of discrepancies that raise questions about when is an error an error.
    • A common abbreviation on a forename. In my 1881 census family Ancestry, FamilySearch and the image itself used the abbreviation. FMP transcript had the full name. I went with the majority.
    • A 1911 entry where a forename had been hyphenated. Ancestry didn't, FMP did. You could just about see it on the page so I've put it in. Would that count as a fail?
    In both cases the household is large so no match lost.
  16. peter

    peter Administrator Staff Member

    That's because you changed the entry rather than confirming that your existing entry was correct.
    In many cases the near match will be with one of your existing contacts - you can check with them. If all else fails you can contact me - that's what I'm here for, to deal with the very rare exceptions that occur so infrequently that personal intervention is the only sensible option.
  17. peter

    peter Administrator Staff Member

    In 1881, on those rare occasions where there is a difference, always go with Ancestry (even if all the other sources disagree) as they are the ONLY site to have the unembellished LDS transcription of this census. The only other sources of the original LDS transcripts are microfilms held by some record offices and family history societies, and CD ROMs.

    As far as 1911 is concerned it doesn't matter what has been transcribed - all that matters is what the householder wrote*. If you and one of your cousins interpret the handwriting differently it will show as a near match.

    * you can refer to the householder's signature to clarify what was intended
  18. Pauline

    Pauline LostCousins Superstar

    The person who completed the form wasn't always the head of the household, and sometimes the spelling in the signature differs from what was written above. In these cases I've always gone with how the name was spelt in the household list rather than in the signature. I've only ever used the signature spelling if the writing above is open to interpretation.
  19. peter

    peter Administrator Staff Member

    It doesn't matter who completed the form, if the same person signed it your interpretation of the information on the form can take the signature into account. For example, it's very common for someone filling out the form to leave out a letter from the surname.
  20. Pauline

    Pauline LostCousins Superstar

    Taking one of the examples I mentioned above, the form was signed by the head of the household, but he almost certainly wasn't the person who filled out the form as the handwriting is very different. So the forename of the head of the household was spelt Abrham on the form (by someone else) and spelt Abraham in his signature.

    So while it seems pretty likely with this particular example that Abraham was intended, nevertheless transcribing it as Abraham would seem to me to require a degree of interpretation that transcribers are not supposed to use.

Share This Page