Jump to content
Great War Forum

Remembered Today:

Chris_Baker

Who can find the worst indexing at Ancestry.com?

Recommended Posts

Christina

Whats new!! I found my Crawleys on census listed as Shawley and Cranley, My Tuckers listed as Lucker. While census are handed to foreign countries to translate we will continue to get these errors. I am sure plenty would offer to translate for free if given choice of area/name to translate. At least we would be familiar with English names and places.

Share this post


Link to post
Share on other sites
ollydot

A word of warning re the free trial period with Ancestry, cancel well before the 14 days are up, it can take 2 days to cancel or it goes straight to your card for a years membership. Also the free trial is limited in access. A lot of the people who transcribe the records are volunteers as far as I'm aware like familysearch.org, so you are at the mercy of individual volunteers and how diligent they are.

You think the English records are bad I have found some Irish records with the placename given as ..... Ireland, England and one notable one ..... Ireland, Isle of Man.

And here's me thinking the Americans all came from Ireland. When they come back here they have the accent and the address of the cottage their family was thrown out of before they starved at the roadside.... but they really just want to be like the rest of us Irish,........... all twinkly eyed ******* with a pig under our arm.

Seriously though take car aboutr relying on Ancestry for accurate records I always work on the principle that I need at least 2 different sources before I accept a record as accurate, been burnt too many times.

Share this post


Link to post
Share on other sites
Max (UK)
Seriously though take car aboutr relying on Ancestry for accurate records I always work on the principle that I need at least 2 different sources before I accept a record as accurate, been burnt too many times.

I don't get this. Ancestry generally allows you to see the Original document - there IS NO BETTER source.

I could understand what you just said if the originals weren't available, but they are :huh:

Share this post


Link to post
Share on other sites
Alan_J

Not only indexing, but missing pages. I researched a soldier and found two pages relating to him on Ancestry's version of WO/364. At Kew, I found a further five pages, with a great deal of extra information.

Going back to the question about a review process, public sector bodies are required to demonstrate impartiality in the tendering process and 'value for money' in the bid accepted, and many of the review processes are geared towards this. This is not at all the same as a consultation and certainly not a public one. The people who check fairness and value are other Government departments and perhaps the NAO.

I think the basic messge is two-fold - first don't rely on the indexing as someone said early on, and secondly don't rely on what you do find (if you do) being all there is to find. Check the records at Kew if you want to be sure.

Alan

Share this post


Link to post
Share on other sites
Stephen Nulty

I don't see what all the fuss is about.

As a long standing member of Ancestry, I am well used to transcription errors. They exists - get over it!

The way I see it, for no extra charge above what I have been paying Ancestry for several years, I have recently been given access to Pension records and now Burnt records that previously I would have had to travel to Kew to access. The price I pay is that I have to be creative when searching.

Fantastic news, I would say.

What does it matter if the government of the day allow a private company to digitise these records and then charge for access to recoup their outlay? The originals are still at Kew for those who prefer this route. Nothing has been taken away from you, so what's the problem?

Share this post


Link to post
Share on other sites
saberhagen427

These errors are unavoidable if they're going to do it as cheaply as possible. Even if they used people with better skills and knowledge, it's impossible to eliminate errors completely.

What I'd like to see is a better system for correcting them. Suppose users could actually edit the index data themselves, and changes went live after being approved by a moderator within a day or two. That would be really good wouldn't it? If they gave us the opportunity to be more involved in correcting and updating the index, we'd take it rather than moaning about them on forums.

Share this post


Link to post
Share on other sites
stevew
May I ask if anybody knows what is happening to the Family Records Centre @ Islington as I understand it is closing ?

Where will I have to go if I want to visit in future (other than the web)....Is it going to Kew ?

It's closed already, the records have gone to Christchurch, but you can't view them. I am not sure what the score is with them at the moment but I thought they were going on line, but the work isn't completed yet

Share this post


Link to post
Share on other sites
shinglma
The way I see it, for no extra charge above what I have been paying Ancestry for several years, I have recently been given access to Pension records and now Burnt records that previously I would have had to travel to Kew to access. The price I pay is that I have to be creative when searching.

Fantastic news, I would say.

You will be delighted to discover then that the online records are missing pages which you can only view on the microfilms at Kew.

I've no objection to them putting them online providing they do put them online completely. :blink:

Regards

Mike S

Share this post


Link to post
Share on other sites
shinglma
It's closed already, the records have gone to Christchurch, but you can't view them. I am not sure what the score is with them at the moment but I thought they were going on line, but the work isn't completed yet

Its true the registers have gone but the FRC is still open in so far as the first floor computer access will remain until early 2008. The shop also remains open.

Useful if one wants to look at the online records free of charge.

Regards

Mike S

Share this post


Link to post
Share on other sites
Greyhound
What I'd like to see is a better system for correcting them. Suppose users could actually edit the index data themselves, and changes went live after being approved by a moderator within a day or two. That would be really good wouldn't it? If they gave us the opportunity to be more involved in correcting and updating the index, we'd take it rather than moaning about them on forums.

Yes! Yes! Yes!

What a good point, I wish I'd thought of saying that when I recently completed an Ancestry questionnaire.

Share this post


Link to post
Share on other sites
Chris_Baker

Anyone searching for soldiers from Axminster in Devon is going to have a hard time.

post-1-1197190609.jpg

Might as well give up and have a Derconshire cream tea.

Share this post


Link to post
Share on other sites
saberhagen427

Another possibility is that we could use Your Archives to make our own index. There are already a few transcripts of MICs and PoW reports up there.

Share this post


Link to post
Share on other sites
Jim Clay

I present this snippet of transcriptions from the 1861 census for interest only - no claim that they're the worst!

post-2135-1197377410.jpg

The correct names are generally obvious - although Herman, Middlesex beats me at the mo. :rolleyes:

Jim

(and the surnames look a bit dodgy as well...)

EDIT:

Herman is probably Harrow, see:

post-2135-1197378625.jpg

and the name for this record is probably Tanner, see:

post-2135-1197378687.jpg

I do sympathise with the transcribers - but a total lack of knowledge of UK place names doesn't help... :wacko:

Edited by Jim Clay

Share this post


Link to post
Share on other sites
Sue Light

I would say that Herman = Harrow ?

But quite honestly, it's easy enough to slag off Ancestry and the transcribers, but anyone who spends a lot of time looking at originals knows what an impossible job they can have. And going back to one of Chris's points right at the beginning of the thread, if an entry is incorrectly spelt in the first place - the example he gives of 'Marleybone' is spelt that way in the original - is it the transcriber's job to correct, as well as transcribe? Personally I think not.

Sue

Share this post


Link to post
Share on other sites
MartH

For any efficient quality database to work you need as smallest index as possible, and data to be as correct as possible, also field used for searching should be checked too.

The fault here is not the transcription but a stage in the process was missed, a data quality check stage.

It is quite easy to do, you check this list to make sure only correct characters have been entered, referenced against a standard list, for example place names. A list of words that should not be in a field, which you build your self by doing an alpha list of the field and using ones eyeball to build a list of illegal words, etc.

This is all standard professional quality large database build work, that from the examples shown clearly has not been done.

Clearly an overall approach by Ancestry because doing this with one data source allows you to rapidly build it for other ones, substantially reducing the cost of build and complaints. Likewise the better the cross referencing the data, the higher the likelihood of a search hit, the more people join and use the data and hence the higher the revenue from data already scanned.

Chris is right but it also hurts them

Specialist subject of mine data quality sorry,

Mart

Share this post


Link to post
Share on other sites
Jim Clay

I would say that Herman = Harrow ?

But quite honestly, it's easy enough to slag off Ancestry and the transcribers, but anyone who spends a lot of time looking at originals knows what an impossible job they can have. And going back to one of Chris's points right at the beginning of the thread, if an entry is incorrectly spelt in the first place - the example he gives of 'Marleybone' is spelt that way in the original - is it the transcriber's job to correct, as well as transcribe? Personally I think not.

Sue

I agree, it's all too easy to criticise - the illustrations above give some small indication of the poor originals (or, rather, images of originals) that the transcribers have to work from. And I certainly agree that it is NOT the job of the transcriber to correct 'errors' in the originals. TWYS applies to all of these tasks.

I do think that some more quality control would be useful, especially where, it's obvious, the transcribers have absolutely no knowledge of the UK. But this would add to the cost of putting the data online (and the cost to us, the users) and would inevitably delay the release of the data.

What do we want? Relatively cheap access, and access now - or near as d*mn it? Or do we want high quality, thoroughly checked indexes, with an inevitable increase in cost, and extended delays in release?

I, for one, am grateful to Ancestry for making so much information available at reasonable cost. I wish they could afford to employ an army of qualified UK-immersed transcribers, but I'll put up with the thrill/chore of the hunt, 'cos I know, in honesty, they can't. ;)

Jim

Share this post


Link to post
Share on other sites
Sue Light

Jim

I agree with your thoughts [mutual admiration society? :P ]. And I also feel that data quality checking might have limited success when so many entries are illegible, and where entries such as place of birth often bear no relation to the town or county of census. We may be a geographically literate lot here, but even to gather an army of transcribers in this country, and pay them a pittance, would still throw up a majority [iMHO] who are not deeply knowledgeable about parishes of Britain circa 1900.

I don't like to harp back to 'In my day' [Oh yes I do!] but if you remember the trials of trying to do family history with NO indexing of any sort on any census, and NO computers, and with the nearest access to BMDs on fiche about fifty miles away, I feel privileged to have what we have today, even if imperfect.

Sue

Share this post


Link to post
Share on other sites
MartH

Sue

Your example of knowledge about the Parishes of Britain circa 1900 is exactly the sort of thing Ancestry should have supplied to the transcribers, they could have easily supplied some simple guidelines and a list speeding up the transcription. It's one of the reasons, their knowledge of the data and transcribing it, is why they should have got the contract.

It's standard quality large database build.

Mart

Share this post


Link to post
Share on other sites
Greyhound

I get the impression that the transcribers were paid by the page (rather than by the hour, say), as it often doesn't look as though much time has been spent trying to decipher some of the more unclear entries. For instance, examples of certain letters are usually readily available on the page (S in son, D in daughter etc), yet the transcribers often don't seem to have made the effort to compare them.

It is a pity, but I'm another in the "grateful to have it at all" camp.

Share this post


Link to post
Share on other sites
Sue Light

Mart

But if it's incorrect or illegible in the first place, as so many entries are, both in census returns and other documents, then the best Gazetteer is of extremely limited use. I have some great examples just within my own family documents, which I would upload if only this b****y software would allow it!

Sue

Share this post


Link to post
Share on other sites
MartH

Sue

What I'm saying is that the guidelines should have said:

If illegible - don't make it up we will check.

Some sort of simple Gazetteer, with a more elaborate one in the system.

The move of data from the residence, or birth parish to another field if it is not confirmed by an internal list and logic (including gazetteer), so it still allow's a search, but does not stop a search when using "know correct" search data on the person.

You can identify the majority of incorrect data and effectively label it as such, people would still see it, have quicker searches, and it would benefit ancestry.

Any examples gratefully received, always love them.

Regards

Mart

Share this post


Link to post
Share on other sites
Sue Light

I believe you! I believe you! :o

Let's try again - I know what these are, but I would wager you'd need to be a Dorsetshire buff to work them out [Top two - even I can read New Haven].

Must just add that the census is not for Dorset, but for Teddington, Middlesex, so no clues could be picked up there.

Share this post


Link to post
Share on other sites
saberhagen427
What do we want? Relatively cheap access, and access now - or near as d*mn it? Or do we want high quality, thoroughly checked indexes, with an inevitable increase in cost, and extended delays in release?

False dichotomy. It's good that they've released these records quickly and cheaply, but that doesn't preclude giving users the opportunity to supply corrections, which some of us would be happy to do for no money (after all, if everyone had to be paid to supply historical information this forum wouldn't exist). The mistake that Ancestry, like many other digitization projects, appear to have made is to assume that a digital resource is ever finished. There are always more corrections or improvements to be made. No source is ever likely to be perfect, but we can try to make them better.

Share this post


Link to post
Share on other sites
MartH

Sue,

OK

First one: Purewell

Second one: is much harder can I some back

I cheated I used this link Online Gaz

Mart

Share this post


Link to post
Share on other sites
Sue Light

Mart

The first one is way out I'm afraid! I won't turn it into some gruesome guessing game, but it's a really good example of how things occurred. The first one is, in fact, Puncknowle, which was/still is pronounced Punnell. So Punnell was what the enumerator was told, and that's what he wrote down. The second one, not that far away is Powerstock - told to the enumerator I'm sure with a broad Dorset accent as 'Poorstock' which was often used as an alternative spelling. And I guess that many soldiers were dictating village names at attestation in the same way.

I just regard this whole British indexing thing with some tolerance, and feel that it was then, and still is, mighty difficult to sort out.

Regards ----- Sue

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...