Blog item - Editing Dates and Places on Indexed Records—FamilySearch Update

LegacyUser

Paul said: I have queried a couple of points relating to this article.

Firstly, I have concurred with Brett's post (see https://www.familysearch.org/blog/en/...) that I think the Blog poster has misunderstood the intention of this enhancement.

Secondly, I have requested an answer to the point made by Elizabeth Hinkle - effectively asking if she needs to edit the 44,000+ errors (wrong place name) that have been applied to a particular collection.

Assuming my post gets through the moderation stage, I am not too optimistic of Brett, Elizabeth or me getting a reply, however.

LegacyUser

Paul said: Incidentally, I am perfectly happy for the feature to be used to highlight the correct information, whatever the reason for the error (indexing error, or error in the original document). As long a reason statement is available, let other users make up their minds about what really is / was the correct place, date, spelling, etc.

The important thing to me is to be able to present any reasonable, alternative information. It's just that I thought this was not FamilySearch's idea - which was for the feature to purely address indexing errors. Maybe the developers decided to be flexible about the provisions the feature can offer.

LegacyUser

JimGreene said: Paul, I have replied to Elizabeth and to Brett in the original blog post that you are referencing. To summarize: No you do not need to correct the place in the entire record-set, it is clear that we, FamilySearch, incorrectly categorized data when we posted the record-set. I have let the right person know and FamilySearch will make the macro-change. As Elizabeth did, just let us know when record-set-wide issues exist.
I gave Brett an example where it would be acceptable to correct both the record and the index, after complete verification. Hijacking records is an abuse that should be reported to FamilySearch. We did intent to allow the family to correct both scenarios, and having a discrepancy between various corrections is also perfectly fine. We are not looking for the holy grail of perfect records, if such a thing even exists. We simply want the tree to be a repository of what is known about each person, complementary or contradictory, each has its place as long as the information is attached to the correct person. One piece of contradictory information does not mean it is attached to an incorrect person, nor should it be thrown away. It is all a part of the record of life for that person. I hope I have explained that clearly enough. if not I'm happy to clarify if there are still questions.

LegacyUser

Paul said: Jim

Thank you very much for your response. My pessimism about us receiving a reply came from experience of the blog item authors often writing their piece, but never returning to respond to the resulting queries.

With regards to your suggestion that batch-wide issues can be corrected, I cannot share your confidence these will be addressed. I have reported these through Support and through this medium, too. None, to my knowledge, has ever been corrected. The usual problem is an incorrect parish name being applied and - as in Elizabeth's example - this usually affects thousands of records. If a similar mistake relates to a Catalog entry, there is a direct link to allow it to be reported / corrected. However, unless you know of a dedicated email address (or internal link) for us to report mass indexing errors, I think you will find FamilySearch does not have resources to correct errors of this kind.

Nevertheless, I am very pleased at your interest in this issuem and your specific clarification about the (updated) feature, which should certainly lead to an improvement in the integrity of the data shown against these records. Thanks again.

LegacyUser

Brett said: Jim

That is very interesting ...

I cannot see, any "Reply" from you, to my "Comment", in that original "FamilySearch" 'Blog' Article.

I have just check through my e-mails, I cannot find any "Response" from you, to me, in relation my "Comment", in that "FamilySearch" 'Blog' Article.

Perhaps, I missed it; but, I thinks not.

And, in any case, if I would have received such a response to my "Comment", that it is considered acceptable to "Change" the actual CONTENT of the actual Record, to what the User/Patrons feels its should be, I would have remembered that; as, I would have been; and, in fact, am, "vehemently" OPPOSED to that.

As I stated in the "FamilySearch" 'Blog' Article:
------------------
Brett says:
March 31, 2020 at 6:55 pm
FamilySearch
------------------
Quote:
------------------
This Article SHOULD contain, a ‘Rider’, to the fact that, ANY such “Edits” of “Indexed” Records should ONLY be a correction to an incorrectly indexed record; and, NOT, as “Edit”; as to WHAT the Records should be (in the view of the User/Patron making the correction) – in other words, what “Actually” APPEARS in the “Image”; even, if that (ie. what appears) is WRONG.

Just my thoughts.

Brett
------------------

AND, more importantly ... I stand by that "Comment" ...

I am sorry to contradict.
If you did respond; and, I cannot, either, remember; and/or, find it - I stand corrected.

Brett

.

LegacyUser

Paul said: Brett

Perhaps I should have waited and let Jim respond directly, but I imagine the reason you are not seeing his replies (to you & Elizabeth) is related to the fact that - even though an employee - his comments also have to be vetted (moderated) before publication. Hence the day or so delay before our comments can be seen there. (You will note my post of yesterday cannot be seen, either.)

I am not surprised at your reaction to the "news" that the edit function is to be extended from its original purpose - I expect Tom Huber will be equally dismayed! I'm quite comfortable with this, as the original version is not being changed and I welcome any alternative suggestions affecting accuracy of detail - whether the mistake was made during the indexing process OR in the original document.

As I have stated, other users can see the evidence and decide for themselves what is correct, as nothing is being CHANGED, so the original detail remains (as illustrated in the Blog article).

Anything that will help find my "hidden" LATHAN relatives (currently both recorded AND indexed as everything from LEATHER to SATHAN) is a great enhancement as far as I'm concerned.

LegacyUser

Brett said: Paul

I very am well aware of the vetting (moderation) process, of "Comments", in "FamilySearch" 'Blog' Articles. In fact, I watched/followed my "Comment" through that process in that "FamilySearch" 'Blog' Articles; and, it day not take too long - the next day.

I am sorry; but, x6 Days to have a response from 'Jim', to my "Comment", vetted (moderated), in that "FamilySearch" 'Blog' Article, is far too long.

In such case, for 'Jim' to respond in this Forum, in the manner in which he did, as if we had received his response, is very poor indeed - unacceptable, in fact.

And, in relation to the matter of, it being considered acceptable to "Change" the actual CONTENT of the actual Record, to what the User/Patrons feels it should be, rather than what it ACTUALLY is, I am "vehemently" OPPOSED to that.

I am certainly no Genealogist; but, is that not contrary to Genealogical Practice.

Correcting an "Transaction Error" in "Indexing" of a Record, is one thing, that is acceptable; and, should be done.

Whereas, "Changing" what ACTUALLY appears in a Record, whether correct or not, to what a User/Patron feels should be the case, whether or not that can be supported by/with evidence, that is NOT acceptable; and, should NOT be done.

Prior to the ability to CORRECT "Indexed" Records ...

In my case ...

There have been many, many instances where there was "Transaction Errors" in "Indexed" Records; where, I added a Prefix 'Rider' in Brackets "[ ]" to the "Source" 'Title', to that effect.

And, there have also been many instances where the actual Record was incorrect (eg, phonetic pronunciation), that could be supported by/with evidence; where, I added a Prefix 'Rider' in Brackets "[ ]" to the "Source", to that effect.

I would prefer that a CORRECTION to a "Transaction Error" in an "Indexed" Record appear more PROMINENT; whereas, IF, a "Change" to what a User/Patron feels should be the case, whether or not that can be supported by/with evidence, is permitted; THEN, such to be LESS prominent, even to the extent of being within Brackets "[ ]" beneath the actual Detail in the Record.

I am sorry; but, "Change" to an "Indexed" Record, to what a User/Patron feels should be the case, whether or not that can be supported by/with evidence, is really NOT acceptable, by any stretch of the imagination.

Having a "Change" to an "Indexed" Record, to what a User/Patron feels should be the case, whether or not that can be supported by/with evidence, appear as LESS prominent and within Brackets "[ ]", beneath the actual Detail in the Record, would possibly be passable; but, that is all.

Brett

.

LegacyUser

brentsweeny said: Jim, I'm happy to hear that there is some hope of getting macro errors in indexing fixed, but not sure how to take advantage of that mechanism. I opened a ticket a few years ago about how a whole tranche of 3-image Indiana marriage records were characterized as single-image, so the resulting indexing was horribly incomplete and fragmented (as you might expect when you see a page with only the groom's name and some of his information and the bride's father's name, for example but nothing else about the bride!). Is there any hope of getting that set of records made right? Please let me know how I should report it. thanks. brent

LegacyUser

JimGreene said: brentsweeny, there is not a formal process set up for doing these corrections to record-sets, it is still pretty rare and so we handle then ad hoc. The best way is actually to inform us via this GetSatisfaction/Feedback forum. We are watching and will get it to the right people. If we don't then we also don't mind a gentle reminder either. For the instance you site above with Indiana marriages, open a new feedback with the details on the records and I will forward it to the right folks. I am not sure of the outcome after this long, but it is always worth a try!

LegacyUser

JimGreene said: Come on Brett, this is a community forum and we need to treat each other better than that! I do not want to engage in a **** for tat, however, you need to know the facts: I do not regularly read blog posts. In fact, I do it rarely. I monitor this forum and others monitor the blog--apart from our day jobs. I only saw the referenced blog post because in this forum Paul Wrightson drew my attention to it. When he did I replied immediately on the blog and on this forum. I expected a delay in moderation for my blog comments, but I assumed that it would come quickly--perhaps a bad assumption on my part. Nevertheless, I also assumed that someone familiar with the blog would know that if my comments were not immediately visible that they would be as soon as moderated.

It sounds like we are just going to have to agree to disagree on our preferences with regard to editing indexed records. Thank you for your input and opinion.

Now, in the spirit of trying to be kind and friendly and inviting in this community I will say no more, other than to ask you to please not judge or ridicule others, their motives or actions. This is not a forum for that. Here we give product feedback and suggestions/opinions. Thank you! I do value your contributions and ideas and the time and effort that you spend on these many forums that you contribute to. If you would prefer, in the future I will offer my comments to you via email, just let me know.

LegacyUser

JimGreene said: Paul Wrightson, thank you for your replies and kind responses. As I explained to brentsweeney reporting details of record-sets with widespread errors can be done here in these feedback forums and one of us who monitor them will forward details to the correct folks to repair. I have already heard back that Elizabeth's issue is being worked on now.
Regards

LegacyUser

Brett said: Jim

'Yes', you are correct, "... this is a community forum ...".
Whereas, you are incorrect, I was NOT treating you badly, nor was I engaging "... in a '**** for tat' ...".

I just took again, moments ago, your "Responses" that you indicated that you made to us in that "FamilySearch" 'Blog' Article have STILL not been posted, even if that was ONLY a Day or TWO ago.

So, I suggest that my response was not the gibe that you imply, that I was treating you badly.

And, 'Yes', hindsight is great; whereas, assumptions are not.

Agreeing to disagree between participants in this Forum is one thing; but, with "FamilySearch" appearing to consider it acceptable to "Change" the actual CONTENT of the actual Record, to what the User/Patron feels it should be, rather than what it ACTUALLY is; even if, that could be supported by/with evidence, is just wrong - so, lets all go ahead and just "Change" FACT for what we want/believe it to be.

The reference/implication that I am judging or ridiculing others, their motives or actions, is so WRONG; and, so over the top that, I am actually offended and appalled with such an accusation/implication.

It would appear to me that you do not actually value the Help that I give to others in this Forum or elsewhere; nor, my opinions or suggestions - that is fine.

I am disappointed with your response to me.

And, the implication in the following reply, referencing 'Paul's' "kindness", is even more insulting towards me - pointedly so.

This is not '**** for tat', I will not stand by and be maligned in this matter, to be made out to be in the wrong or the bad person, when I am not and was not.

Generally, I admit when I am wrong; and, I am quite often wrong.

Please do not respond - you have had your say and I mine - let 'bygones be bygones'.

Enough said.

Brett

.

LegacyUser

Juli said: Jim, you're completely wrong about macro-errors being rare. The database is rife with them, and reporting the errors using the available channels has no effect beyond raising my blood pressure. (No matter how I phrase them, my messages are willfully misinterpreted as a desire for index correction.)

For the Hungarian civil registrations (indexing ongoing), any time a film has multiple places on it, the first one is assigned to everything that is indexed from that film. This leads to index entries in the wrong county and other such ridiculousness.

Some examples from "My Cases":
https://www.familysearch.org/ark:/619... "Gibárt, Abaúj-Torna" should actually be Görömböly, Borsod
https://familysearch.org/ark:/61903/1... "Felsőnyárád, Borsod" should be Felsővadász, Abaúj-Torna

LegacyUser

Paul said: Agreed - multiple items on the one microfilm has been the biggest problem for this (common) problem. Two of my examples relate to parishes in Norfolk and Northumberland. I mention them here from time to time but have given-up on them ever being corrected.

Not unique to FamilySearch (I've find the same problem of misidentifying parishes in Find My Past) but other sites do seem to have a proper reporting / correcting process.

LegacyUser

Carolyn Wheeler said: I also agree with Juli and Paul. By my own experience many of my parishes in
southern Germany are often misidentified when indexed. It is a source of frustration because there is no way for me to correct the errors. This is definitely not a rare problem, but instead is quite common.

LegacyUser

Adrian Bruce said: "Agreed - multiple items on the one microfilm ..."
Also agreed by me.

"... Not unique to FamilySearch (I've find the same problem of misidentifying parishes in Find My Past)"
I have seen it in FMP's Cheshire PRs where entries for Wrenbury parish are indexed as Wilmslow parish because that's the parish at the front of the film. (Or, of course, it could be vice versa). Mind you... Who supplied FMP with the index for their Cheshire parishes? (Insert evil grin....)

Not blaming you Jim but, shall we say, it is an ongoing saga with which some of us are all too familiar.

LegacyUser

JimGreene said: Juli, Paul, Carolyn and Adrian, thanks for your input. I get it I am not altogether sure that the problem Elizabeth encountered that started this thread is the same as the ones you are all describing, but I will promise you that I will talk to those who know and see what they tell me, then I will report back to you here. I suspect that I am going to have to have more details on each of the problems that you encountered if it is decided that it is something that we can fix as easily as we could fix Elizabeth's. But stay tuned, it may take a bit, say a week or two, but I will get back. Gentle reminders always welcome.

LegacyUser

JimGreene said: I want to reply to Elizabeth (blog), Juli, Paul, Adrian and Carolyn. Thank you all for not only replying to let me know that it was not a "rare" occurrence, but for also giving examples. I was able to get these examples to the correct people. It started a review process and an audit process that showed us we had a deficiency in our current methodologies and processes for fixing these record-wide or macro issues. From this realization we are defining changes and working toward a methodology that is much more responsive and timely in making corrections. Thank you thank you!
For now continue to report these issues here on this forum, and we who monitor this forum will continue to get these to the right people. Eventually we will announce a formalized process. Again, without you all taking the time to let us know things would not have changed.

LegacyUser

Adrian Bruce said: Thanks to you, Jim, for triggering that review process. Contrary to what some might think, all many of us want to do here is to get bad processes changed. Many of us have been there in our professional lives - any job that I never made a mistake on, probably wasn't worth doing...

Best of luck with your future methods.

LegacyUser

Paul said: Yes, very grateful to you for your efforts relating to this issue, Jim.

LegacyUser

David Newton said: That is a very positive development indeed.

To kickstart the process of reporting we have the Hampshire parish register records where the places are of the format Place, England rather than the format Place, Hampshire, England, (United Kingdom) with the terminal United Kingdom being dependent on the year of the event of course. We also have the Staffordshire records of the format Place, West Midlands, England, United Kingdom where West Midlands did not come into existence as an actual county until 1973, long after the date of most of the events purportedly taking place thete.

Both these large sets of problems are in the fairly recently released Jampshire and Staffordshire specific databases, not the old IGI-derived database. The Hampshire problems are in the non-bishop's transcript database.

LegacyUser

JimGreene said: Thanks David. I have forwarded this on.

LegacyUser

Jeff Wiseman said: It does my heart good to hear when contributions from the forum can result in significant process changes.

LegacyUser

Adrian Bruce said: Jim - here's an interesting one from a process viewpoint. I'm really offering it up for you to illustrate what is probably one of the more complex issues I've found.

https://www.familysearch.org/ark:/619...
is indexed as the burial of a female named Johnson, wife of William Johnson. Burial is ostensibly on 27 Apr 1828 at Church Minshull, Cheshire, England.

Because I have access to the corresponding images on FindMyPast, I can see what this so-called burial actually is. At the top of the page (and I am assuming that FS has access to the same image) is the heading "Churchings in 1828". The entry for Mrs Johnson is at the top and reads
"April 27, William Johnson wife, Minshull"

Now the opposite page is headed "Burials 1828", so this has been interpreted as the burial of William Johnson's wife on that date (I assume Minshull is the residence). However, an experienced researcher should start worrying - if this is a page of burials, why are they all wives? Why are there no first names for the wives? Why no ages (the burials on the LHS do have ages)?

The explanation is in the heading "Churchings". The ceremony of churching was carried out on mothers who'd recently given birth and was intended to bless them - it was derived from purification rituals. (Gross over-simplification going on here). These are very much not burials.

Now, I can't blame anyone for not knowing what churching was - not sure I would have done, were it not for a story my mother told me about my Nana insisting that Mum be churched after my birth. I don't suggest that indexers or arbitrators should have known about churching beforehand - it's so rare to see it written down that I would never have expected it. But I am concerned that (apparently) indexers and / or arbitrators together didn't react to the anomalies (all wives, etc.) or the heading describing this unknown ceremony, and seek guidance how it should be recorded (no, I don't know how either!).

To me, personally, there are two major points here:
1) Index project training should include not just the facts of what to do, but also techniques of how to recognise the anomalies that say that the ground just shifted beneath your feet and then how to deal with it. And no, not everyone will be able to recognise the anomalies.
2) Correcting the rest of the entries on that film where Churchings have been Indexed as Burials - I have no idea how many there are.

Hope this is useful Jim.

LegacyUser

JimGreene said: Sorry Adrian for the long delay, I was hoping to be able to give you a better response than the one I am going to give. To be honest I am scratching my head on this one too. As I have discussed this with others we feel that something like this may be rare. I know the last time I said that I was immediately proven wrong, but hey, that is why we talk, right? Under the premise of rarity, our answer to this is that over time, a piece at a time, users will have the ability to change just about anything in the indexed record. Rather than try to code to the anomaly or teach a level of indexing knowledge that, quite frankly, the majority of indexers don't want--they feel we are too complex as it is--it would be left to the super indexers to correct something like this, with the new system. And they would be able to we believe. It sounds like this is just a page or batch of data, unless I am wrong. I am willing to be proven wrong and will gladly take more data back to the decision makers. I have sent this instance off to see if there is a mass correction for this record that can be done. Thank you for a great catch and a wonderful explanation!

LegacyUser

Carolyn Wheeler said: Thank you, Jim, for your very thoughtful, detailed, and encouraging responses. I must say that the inability to correct place names has been so frustrating that I just gave up on it and quit tracking the misidentified towns and parishes, but now that I know there is hope I will note them as I come across them.

I do have two other questions about correcting indexed records. I hope this is the place to ask these questions as I don’t mean to hijack this thread.

1) I love that we can correct misindexed given and surnames. However, there is a problem with persons who only have indexed given names with no indexed surnames. Recently I have come across a number of records (usually wives) with only indexed given names. When I read the original images the surnames are plainly stated, but I cannot add the missing surnames because no surname field is available. I wish that I could add the missing surnames, but cannot do that under the current setup. Is it possible that this situation could be fixed so that missing surnames can be added to indexed records?

2) Why do some indexed records where the wife’s maiden name is clearly given in the original image record her married name for her surname? This seems to break the rule of recording women by their maiden names. In each of these cases her maiden name is indexed as an alias instead of as her surname thereby listing two surnames for her. This is confusing. I understand what “alias” means In these cases because I can read the original records, but persons who cannot are left to wonder what “alias” means. So again, why are maiden names recorded as aliases instead of being recorded in the surname field?

Thank you for answering these questions. Thank you for participating in this forum.

LegacyUser

Adrian Bruce said: Jim - thanks for your reply.

1) I would (mostly) agree that trying to get the indexers to deal with stuff like the Churchings wouldn't be practical. They're obscure and unexpected - and (until I found them) were at the very least "Unknowns" to me and possibly even "Unknown Unknowns". My only caveat that leads me to say "mostly agree" is that I think there should be (and, I don't know, maybe there already is) a mechanism for the ordinary indexers to raise issues with someone (the super-indexers?) to say "This looks a bit weird to me - are we doing this right?" (This isn't the same sort of thing as asking what data goes in what column in a specific indexing project - though the queries may go to the same people.)

I'd like to see a clear path for "This looks a bit weird to me ..." queries built into any indexing project - if it isn't already. And indexers should be encouraged to raise such queries, i.e. the path mustn't be hidden in the small print.

2) I presume that "super-indexers" are (or will be?) available during an indexing project?

3) If the super-indexers can somehow deal with things like this during the indexing phase (i.e. preferably before going on general release), then that sounds like the best possible way of dealing with weird stuff like this. Suggestion: Given that we'd really like these things not to be indexed as burials, could there be a generic, user-defined / custom event so that incorrectly attributed burials (etc) could be turned into some sort of custom event? No idea what impact that would have on searching of historical record indexes... :-(

4) Volume of Churching data? I've no idea - Church Minshull isn't a parish I'm familiar with, even though it's adjacent to some of my ancestral parishes. However, I have only just realised that the image in question comes from an item labelled "Rough Register" - 45 images worth for 1814-1835. The references to "Churched" actually start on the very first page, where they are mixed in with baptisms, etc - in fact one or two appear to be part of the same entry( e.g. "X child of Y & Z baptised ddmmyy, mother churched"), which is logical since both Churching and Baptism should closely follow the birth. The final page of Churchings (separate, unlike the first entries) covers 1835 - so there are at least 22 years with Churchings but only 45 images max. No idea if there are any entries in any other registers.

LegacyUser

brentsweeny said: Jim, since you've asked for more details about the Indiana marriage-record problem I referred to earlier, and since I ran across another example of the problem just now, I'll provide some more detail.
I'm looking at a 1905 marriage record from Warrick county, Indiana, part of the collection 'Indiana Marriages, 1811-2007':
https://www.familysearch.org/ark:/619... see images 109-111:
Like other records in its set, it was indexed as a ONE-image record, though the records all have THREE images (20 columns with a foldover middle section) for several years and counties--though I'm not sure which ones at the moment, but I see it often-- The first image is columns 1-7 (groom and part of his family info, with part of columns 19-20 visible on the side), the 2nd image continues with cols 8-13, ore of the groom's info and gives bride name and some of her family, and the 3rd image concludes the bride's info and the marriage details. So, as often happens, when a single image is indexed it results in a groom's name and no bride. other indexing efforts, if they saw only image 2, would get groom's and bride's names, but only some of their parents and no marriage place, etc. They're wonderfully rich records but with the single-image instruction very fragmented.
I have seen some recent Indiana-marriage indexing batches with instruction to view all 20 columns; is the whole set of these 3-image records being redone--probably the only way to fix this problem--or is that just a coincidence? Is this the right place to report that problem? How can I help further?

LegacyUser

JimGreene said: brentsweeny, The answer I got back is that we will be re-indexing these records that cross 3 images. It has already started.
I hope that answers your questions. Basically, we decided to start over and do it right this time.

LegacyUser

JimGreene said: Thank you for your kind words Carolyn. I want to give you hope, but cannot give you dates, for a number of reasons--not the least of which they change so much that I would just end up not telling you the truth!
1) It is in our plan to have an indexing product where you can go in and index more of the record than was originally captured (we only identify some of the fields to be indexed in order to simplify and expedite) on the first pass.
2) Here is what our team replied: "This second item needs more details to give a better explanation. Some collections come from a partner or vendor with an alias field (I specifically recall the German Lutheran records we got from Ancestry with alias and this was how they did it). " FamilySearch usually does not create Alias fields, so I am personally thinking that these were records that were originally indexed by a partner.

I hope that gives you something to hope for and go on!!
Jim

LegacyUser

brentsweeny said: thanks very much for following up, Jim--I think that a re-do is the only way to get it right, and that explains what I saw--sorry it came to that, though.