View Full Version : Ancestry's GRO index issues
I just thought I'd remind people again about the issues on Ancestry regarding their GRO indexes.
George Edwin Winsett is someone being researched on another thread and we were looking for a death entry. We found a death on Ancestry in 1914 which was a generation too late for the person we were after. Here's what happened next..........
I've just realised there is no George Edwin Winsett who died in 1914! Here's the Ancestry entry:
Name: George Edwin Winsett
Estimated Birth Year: abt 1838
Date of Registration: Oct-Nov-Dec 1914
Age at Death: 76
Registration District: Pancras
Inferred County: London
Volume: 1b
Page: 64
and here's the FreeBMD version (this isn't showing on Ancestry's database):
Deaths Dec 1882
WINSETT George Edwin 76 Pancras 1b 64
Note the Q, Volume and Page numbers are the same and the big give away, Ancestry state "Image not available" for the 1914 entry, though the button to order a copy of it for a minimum of £22.99 (:eek:) still appears (I wonder if they would then send the 1882 cert or if the order would be rejected and who would end up with the £22.99??!)
I have sent emails to Ancestry before about this issue, but never received more than an automated reply.
In this case the name was an uncommon one, but it still took a couple of days for me to notice the Ancestry record was a false one!
How?
I can understand typos, transpositions, but the rest of the batch seem okay. And 1882 and 1914 cannot be mistaken for each other..... Is this a computer glitch? The years are 32 apart. A 1 corrupted to a 0 (or vice versa?)
I just went on Ancestry and randomly looked at some deaths from 1914 and 1915. I picked out a few with No Image Available and they all appeared on FreeBMD for other years, but not 1882 particularly - just completely random years. The least misplaced entry I came across was from 1908.
If you do a search without a name on Ancestry's FreeBMD records the first stuff to be presented are all the entries with no surname or just "Male" or "Female" in the index. Quite a large percentage of these had no image, but again they were plucked from random years.
As Ancestry don't seem to care about this issue, I wonder if FreeBMD care, given there name is part of the title of the database. Perhaps Ancestry and FreeBMD are at loggerheads over whose fault it is?!
In the meantime I will try to remember that if I do a search on Ancestry I make a mental note as to whether there is an image available!!
I wonder whether the problem entries are those (or some of those) added to Ancestry's database after the main transfer of FreeBMD entries had happened? That would account for most of the known problems occurring in the later period of the timeframe as much of the earlier stuff was around 100% complete when it was handed over (I think!).
Yes it happened to me - I found the year was wrong - can't remember who it was now.
There are some threads on Ancestry message boards suggesting the problem stems from corrections made to FreeBMD after the data had been handed to Ancestry not being passed on. I doubt that the reason though, as most corrections are just to page numbers or reg district names or incorrect transcription of a name, not entries misplaced by years/decades.
One thread was posted Aug last year. The poster was trying to get Ancestry to correct her mother's birth reg which appeared correctly on FreeBMD in Q3 1914 but Q2 1915 on Ancestry! I see she was unsuccessful in her quest as Ancestry's version is still wrong!
Remembered reading something about this on GF. After a little searching, it was an article from Lost Cousins. This is what was posted:
According a survey of forum members my article about Ancestry's data errors was by far the most popular of the last newsletter. Whilst it was fairly easy to identify incorrect entries that didn't belong in 1872, the real problem was the fact that the marriages incorrectly ascribed to 1872 wouldn't show up in a search of the years to which they actually belonged.
It's heartening to see that as direct result of that article Ancestry have removed almost three-quarters of a million incorrect marriage entries from 1872! I wonder how many 'brick walls' will now come tumbling down?
I haven't been able to check whether they've all been allocated to the correct years, but the entry that led me to the discovery - the misplaced marriage of Solomon Surch - has certainly found its way to 1887, where it belongs. One thing is certainly missing, though - a 'thank you' note from Ancestry - but then again, I didn't do it for them, I did it for YOU!
Unfortunately those aren't the only errors in Ancestry's GRO indexes - following my article, Angela pointed out that there are far too many births in 1915, and too many deaths in 1914. In the 78th Annual Report of the Registrar General he stated that:
"The births registered in the year 1915 numbered 814,614; of these 778,369 were legitimate and 36,245 illegitimate."
Ancestry show 1,697,117 births for the year 1915, which suggests that they have incorrectly recorded over 880,000 birth entries. Similarly in 1914, the Registrar General reported 516,742 deaths, but Ancestry list 820,533 deaths as being registered in that year, over 300,000 more.
For example, Harriet Calver (no relation so far as I know) died in Hendon in 1884 aged 70, but at Ancestry her death is shown as 1914 (with an implied birth year of 1844, also 30 years out).
These are all simple checks that anyone could have carried out, yet Ancestry - the biggest genealogy company that the world has ever known - let nearly 2 million errors creep into their records. It's not as if they are transcription errors - the data was provided by FreeBMD, who show the entries correctly in their own online indexes.
Do you have a date for that article, Shona?
I've just found a thread I posted here in July 2011. At that time I hadn't realised the extent of the problem, just assuming a few entries were misplaced.
Do you have a date for that article, Shona?
23 May 2013
Olde Crone
26-08-13, 10:14
I remember reading that article which Shona quotes and I think there has been a follow-up article too. Very recently (otherwise I wouldn't be able to remember it, lol!).
OC
tenterfieldjulie
26-08-13, 10:16
Wow what a lot of misinformation .. I knew there were lots of errors with names, but not with dates, that is a real worry. It certainly doesn't give you confidence in Ancestry.com, especially if they don't reply to your concerns, or correct something when you prove they are wrong. Julie
I reported the "1915 problem" to Ancestry a couple of months ago before I read the Lost Cousin article that Shona mentioned. I was trying to fill in some missing deaths in my tree and noticed the pattern. I gave some specific examples where the 1915 entries didn't have an associated image, because they were incorrect.
I got a polite but firm reply that I needed to provide the link to the images which were incorrectly indexed in order for the call to be passed on to the team which could fix it. My tongue-in-cheek response was to send them a link to the image on Freebmd (where it was correctly indexed). No reply to that, other than an automated survey on how my call had been dealt with !!!
My tongue-in-cheek response was to send them a link to the image on Freebmd (where it was correctly indexed). No reply to that!
I can confirm that it makes no difference if you send them a link to the Ancestry equivalent page, rather than the FreeBMD one, as they still take no notice!
I think the irritating thing about this is the scale of the problem (and of course the lack of response). It's easy to understand how human error can account for the odd missing census image or a few names at the end of a page not added to an index. These things are hard to indentify and correct individually.
But the best part of a million incorrect index entries is a data load problem and whatever batch process they used to get the entries in there in the first place could be used to correct them in one go. If only they recognise there's a problem in the first place.
I have read (and I know Merry has too) a book called A Comedy of Errors about the errors the GRO made with their handwritten system, but it pales into insignificance against this.
Apart from publicising it as widely as possible, I can see no way of helping the ordinary researcher. Because they will be the poor fools who pay over the odds for their certificates, rather than checking the scan and then ordering in person. The only sop is that they won't find the entry they want, and they probably don't care about death certificates anyway.
tenterfieldjulie
26-08-13, 11:19
You made me go looking for the NSW & ACT State FH Conference which is in to be held next month in Canberra. Two other State Conferences have had speakers from Ancestry but this year they are not being advertised but Find My Past is ... not sure of the significance. Julie
I was invited to the launch of the online Surrey records.
The girl put up by Ancestry to explain what they'd done was too nice and innocent to hurl brickbats at (and it was reprehensible of her superiors to make her show a slide relating to Samuel Pepys without explaining the pronunciation) but you could hear the defensiveness when I queried how the correcting process worked. It sounded like inch by painful inch and that really the request should come not from Joe Public, but from the organisation providing the data. (Surrey has some baptisms entered as marriages)
IThe only sop is that they won't find the entry they want, and they probably don't care about death certificates anyway.
It's not just deaths though. This is part of my 2011 post regarding the same problem with a birth (and I know I've seen others):
This one appears on FreeBMD as 1903 (correct), but on Ancestry as 1915(same page number etc):
Births Dec 1903
BILLYARD Gladys Rotherham 9c 743
I don't know if the issue with births is as bad as the deaths, but wouldn't be surprised. I can't remember if I've found any misplaced marriages.
tenterfieldjulie
26-08-13, 11:51
Merry with the Wimsett/Winsett name being quite rare, if you were aware like you, you would notice the mistake, but with common names, you would just think you had the wrong person if an age was added giving year of birth, but how many people would compare the reg. nos..... not fair to the inexperienced .. Julie
I agree, Julie.
I'm surprised FreeBMD haven't managed to get Ancestry to do something, as on the face of it people might imagine it's FreeBMD who have made a massive errors as the databases have their name in the title.
I agree, Julie.
I'm surprised FreeBMD haven't managed to get Ancestry to do something, as on the face of it people might imagine it's FreeBMD who have made a massive errors as the databases have their name in the title.
Knowing the powerlessness of data providers who were perfectly happy (at least in public) with the terms and conditions before they sold their souls, I regret to say that I am not.
tenterfieldjulie
26-08-13, 12:24
What is the purpose of putting free information on a paid site? !!!
No information is "free". We are dependent on the generosity of the administrators of this site so we can continue to post.
Besides the work of volunteers to do the indexing, there are all sorts of running costs. Currently, the bunker is the only sponsor of Freebmd, but I remember other sponsors in the past. Selling a database is a way to continue to keep it free.
tenterfieldjulie
26-08-13, 12:41
Thanks Phoenix. I certainly appreciate the time and effort of volunteers and sponsors. It is such a shame that commercial entities get greedy and spoil it. When I spoke about errors in one of the paid site to one of their reps., they said it was a rush to get it out the public and they relied on the public to tell them of the errors ...obviously not true if when they are told they won't do anything about it. Julie
kiterunner
27-08-13, 09:06
What is the purpose of putting free information on a paid site? !!!
You don't have to have a paid sub to search and view the FreeBMD databases on ancestry, only to view the image for an entry (if there is one!) One of the reasons why there is a copy of the FreeBMD databases on ancestry is because FreeBMD's own servers couldn't cope with the high demand from researchers. But it is a shame that ancestry's copy has mistakes.
Incidentally, the number of, say, deaths registered in a year would not be expected to match the number of index entries exactly because some registrations have more than one index entry pointing to them, but of course that couldn't account for such huge discrepancies.
vBulletin® v3.8.7 PL3, Copyright ©2000-2024, Jelsoft Enterprises Ltd.