Indexed records should contain detail from the original source and not be based on supposition.

LegacyUser

Paul said: I came across the indexed record shown below earlier today. It includes a "Race" field, where the individual is stated to be "White". This is a quite fair assumption, given the name, location and date the record relates to. However, the parish register would certainly not have recorded such detail, so this should not have been added.



Also, about 80-90% of christening records I come across (relating to England) have incorrect information in the "Residence" field. The name of the parish in which the event took place is nearly always repeated in this field, even though - in the original record - it is either not shown at all, is shown as a minor place within the parish, or even a location many miles away.

In future, indexing procedures / instructions - for whatever stage in the process - should be such that they try to ensure only information actually recorded in the original ends up in the FamilySearch record.

LegacyUser

Adrian Bruce said: The "Residence" thing drives me potty. The correct residence is there on the parish register post 1815 and it's usually(?) ignored in favour of the parish. I have no idea where the residence is added - whether in post-index-processing or what...

But the idea of "Race" being recorded on the index for a parish register of baptisms in the UK...??? Where on earth did that come from? How on earth? I'm just fascinated by how one could get that!! I can't think what on earth could be misinterpreted to imply such a thing.

Notice I am making the assumption that it is an error - it's an assumption that I feel I can make, however.

I would be pretty certain that there will be lots of entries across UK parish registers that refer to "blackamoors", "negroes" and other terms that would make us uncomfortable if used today. However, as Paul says, the converse (i.e. no such entry) should not be used to generate a "White" value!

LegacyUser

MaureenE said: The classification of race as White appears in some of the index records for India baptisms, marriages and burials based on records in the British Library, London.
https://www.familysearch.org/search/c...
These index records were produced c 2010.

This is pure FABRICATION, as nowhere in these records does the wording White appear, and there was no heading in the records about race. The current description of the records is "Index based upon data collected by the Genealogical Society of Utah, Salt Lake City" and a previous description was "Index entries derived from digital copies of original and compiled records".

A few of the actual records (images) do make a comment about race such as Eurasian, but I have never seen this description appear in the index records. As such, I relieve that race was never an item to complete in the indexing, and that the item has been superimposed by FamilySearch at a later stage in the processing, whether as an error or a deliberate decision I do not know. I think it is poor if this was an error by FamilySearch, but even worse if it is a deliberate decision to provide information that was not in the original record.

No where is it advised that the indexes contain entries that were not in the original data, so unless you know about this feature, you may be making incorrect assumptions.

LegacyUser

Adrian Bruce said: Worrying.

Incidentally, I have an Indian chap in my desktop database, who was baptized in London in the 1890s, and there's not a mention of his race. His name is a bit of a giveaway but that's all there is to indicate any particular ethnicity. I don't know if that register is in FS but imagine that it was and the entry had a 'White' generated as the default....

LegacyUser

Adrian Bruce said: If "Race" has been concocted for some parish records from India, and it's appeared on at least one index from the UK, I really think that FS needs to follow up on Paul's example and work out how and where that value came from.

Hopefully it's just a one off but it needs to be investigated just in case it gets on further indexing projects.

LegacyUser

Paul said: I found this thread whilst looking for something totally different! Some of the comments are irrelevant to this topic, but others relate to the "Race field" in English records.

https://getsatisfaction.com/familysea...

LegacyUser

Adrian Bruce said: "So many of the England marriages for which your records attempted to ascribe a "race" were just plain wrong, ranging from a bridegroom in 1873 who was labelled as black because your transcriber mis-read his occupation of clerk, to a bride in 1715 being entered as white because she lived in Whitechapel Road!..."

One wonders if similar errors prompted its re-inclusion here. For instance, was the parish priest who carried out the baptism a "Mr White"? (I don't mean, was that the case on Paul's example, rather, was that seen on some other records that prompted the "race" item to be indexed.)

One respondent on that thread requested that the "race" item be put back saying, "There were mixed race marriages as there are today. The RACE field is very important to those of us of color." Sorry but that's completely misinterpreting matters about that data item. The values in the "race" item in that case seem to be partly made up of a number of errors, and we can guarantee that most entries of ethnic minorities made no mention of their ethnicity. In other words, if she relied on that value, she'd lose the ability to trace her ancestry very quickly.

Again, let me emphasise that there will be a number of entries where "race" was indeed noted but in most cases it was not. (Bear in mind that after 1815(?) registers of C of E baptisms were pre-printed so their column content was defined by central authorities. Didn't stop some people writing extra comments in, of course.)

LegacyUser

David Newton said: 1813 was the introduction of the pre-printed register books for christenings and burials.

LegacyUser

Adrian Bruce said: Thanks David - for whatever reason that never sticks in my mind.

LegacyUser

Emily Oldroyd said: I am over the creation of UK indexing projects and help to manage existing collections for the UK on FamilySearch. Hopefully, I can clarify a little here, this particular collection is one of the oldest index collections on FamilySearch. It was created a long time ago from church records from a lot of religions over England all of which contained varying information over the years. It was one of the first collections published to our website and contains a lot of great content as well as a lot of errors. We have not removed this collection from our website because many people have attached the records from it to their tree. We are however working on analyzing the data with the intent to improve it where we can. This is a long process because it is such a large collection and because we don’t want to cause new errors due to additional treatment of the data.
In the case mentioned above, you are right, the race is not mentioned on this particular record for Thomas. I am not sure why it is in the data but like the rest of the collection this will be looked into and eventually sorted out. In regards to the issues with event place, this is a problem that came up with the creation of our new locality tool which is helping us to standardize and hint. We are working on fixing the localities that got mixed up when this tool was implemented. For your information, we generally do not index race or locality in our more recent UK projects. Thanks for the feedback, hopefully this helps. We appreciate your patience.

LegacyUser

Adrian Bruce said: Emily, thanks for the response. It can get frustrating not knowing whether anyone has heard us so at least we know that you have.

LegacyUser

Paul said: Yes, thank you Emily. A more recent problem is the introduction of an "extra" field for Residence. This is usually for the name of the street in the town/city (the other residence detail) in which the person/family lived. If this is carried across in the source attachment process it almost inevitably creates a data problem warning flag.

Although this might have its uses, it is really creating more trouble for users than it is worth. The solution is either to not include the "High Street" information as an "extra" residence or to prevent Residence data from being flagged as not being standardised. I know the latter would not fall under your remit but please take this problem into consideration.

As stated earlier, the residence detail is not an old problem - the problem of the place the event took place being copied into the Residence section (sometime during the indexing process) appears to be ongoing and should be discontinued.

LegacyUser

David Newton said: This particular collection is the IGI-derived one, so it's old, inconsistent and has a huge number of duplicates in it. This and its marriage equivalent are the source of many complaints about "duplicate" sources that pop up on this forum.

The two biggest problems with it are nonsensical/completely wrong parish names and the huge number of duplicates in it. The worst I have come across was the parish of Chellaston in Derbyshire where the same christening appears in the database no less than 5 times!! From the FSFT point of view all are separate sources and so need attaching. In reality they are duplicates in a database which needs a massive cull. Another example of nonsense in there are entries for the parish of "Froyle, Houts". In reality Froyle in Hampshire. I have also comes across entries supposedly for Worcestershire which are actually for Warwick and entries for "West Clandon, Surrey" which are nothing of the sort (although unfortunately I can't remember what the correct parish is supposed to be).

However this is far from the only collection with serious issues, and age of the collection cannot be used as a defence in these other cases. The Staffordshire collection has large numbers of entries from the 19th century showing up with a county of "West Midlands" as the supposed location. Since that metropolitan county council only came into existence in 1973 it is a mystery how on earth this happened. The more recent entries in the Hampshire database (not the bishop's transcripts one) are showing up with supposed places of Location, England. This is very much an issue for places like Bentley where the location name standardises to Bentley, Staffordshire!! Some event locations are also simply showing up as "Hampshire, England", again not particularly helpful. Some events in the Warwickshire database are also simply showing up with locations of "Warwickshire, England".

Those three collections above show that your workflow pipeline has some serious quality assurance problems. These are not the errors of individual indexers. They are consistent, widespread metadata errors for event locations. If this hasn't already been addressed it urgently needs to be. If it isn't addressed we run the risk of millions more transcribed records with seriously defective event location information showing up in the collections. Oh and please make sure that marriage banns are always clearly distinguished from actual marriages in the indexes.

LegacyUser

Tom Huber said: My gripe is over the inclusion of a burial year in the FamilySearch index derived from Find a Grave.

The Find a Grave entries do not include burial dates and as such, the inclusion of the year, especially when the source linker is used to attach the burial to a person's record, should not be added.

I have mentioned this a number of times and the problem has yet to be addressed.

LegacyUser

Paul said: Emily (hoping you are still looking in)

It would be excellent if somebody from "indexing" followed the posts in this forum. There seem to be so many inconsistencies in the different project instructions, so this really needs to be looked at. As do the metadata problems that are common to collections both old and new - including incorrect descriptions: "buried" instead of "died"; "beneficiary" instead of "executor"; using the event location (parish) instead of the (clearly shown) actual residence from the original register, and the many errors that affect EVERY record in certain collections.

LegacyUser

joe martel said: There are some very active groups in communities specifically for indexing and indexing employees monitor those https://community.familysearch.org/s/global-search/indexing?tabset-2064e=1https://community.familysearch.org/s/global-search/indexing?tabset-2064e=1