* 1921 census

[Johnwrosser]

I'm currently researching with this but only started recently.
It seems that the RG15 numbers aren't shown as 1 line and you have to go down the page to find the rest of the numbers but these don't correspond with the numbers on the actual document? Which are correct?

[fhtess65]

Down which page? The transcription?

I'm currently researching with this but only started recently.
It seems that the RG15 numbers aren't shown as 1 line and you have to go down the page to find the rest of the numbers but these don't correspond with the numbers on the actual document? Which are correct?

[tatewise]

I've moved this to the Research Forum as it is not an FH General Usage topic.

[Gowermick]

If you're refering to the images, you need to see /downlaod all three!
1. Actual schedule, with details of people (Schedule Number shown Top Right as per 1911)
2 Associated Address sheet, to get Address as well as RD, SD & ED references (Similar to 1911 address sheet)
3. Cover sheet to get RG15 & Piece number, with repeat of RD (Name & Number), SD (Name & Number) & ED (Number only)

I think most people forget to download the other sheets

[Johnwrosser]

What 3 sheets? I only see 2, the initial page showing the family then the actual document the 1st page was transcribed from.
I can't understand how they've got different numbers on the original and the transcription?
Never mind the fact that whoever transcribed them must have done it in a rush as some don't match.

[ADC65]

If you are looking at the transcription, the details are as highlighted below:

Screenshot 2022-11-29 161144.jpg (47.88 KiB) Viewed 2987 times

If you don't have access to the transcription, the details are shown in various places. The schedule number is in the top right hand of the schedule:

Screenshot 2022-11-29 161221.jpg (17.27 KiB) Viewed 2987 times

Whereas most of the other details are on the next image (the reverse of the schedule):

Screenshot 2022-11-29 161248.jpg (16.76 KiB) Viewed 2987 times

The piece number is easiest to obtain by looking at the filename of the image you have downloaded - it will the five-digit number immediately after "RG15" - for example:

Code: Select all

GBC_1921_RG15_26167_0444

The piece number is 26167.

[ADC65]

Never mind the fact that whoever transcribed them must have done it in a rush as some don't match.

I would never trust the transcription unless I have no other choice. In my experience, I haven't had too much trouble with the details you are talking about, but the transcription of names, places and occupations (even where the handwriting is quite clear) is appalling.

[Johnwrosser]

Thank you Adrian, very helpful.

[Gowermick]

What 3 sheets? I only see 2, the initial page showing the family then the actual document the 1st page was transcribed from.
I can't understand how they've got different numbers on the original and the transcription?
Never mind the fact that whoever transcribed them must have done it in a rush as some don't match.

The third sheet is no 1 in the series. So if your looking at sheets 59&60 for your household, look also at sheet 1, which is the cover sheet for the whole series, and contains RG15 + Piece number

[Gowermick]

Never mind the fact that whoever transcribed them must have done it in a rush as some don't match.

I would never trust the transcription unless I have no other choice. In my experience, I haven't had too much trouble with the details you are talking about, but the transcription of names, places and occupations (even where the handwriting is quite clear) is appalling.

I agree, my 3’ggmother was transcribed as Jennings Korberry, when it actually said Jemima Rothery

[fhtess65]

I concur...I gather part of the problem was the extreme privacy restrictions which meant the indexers were reading one line at a time, so if unsure they couldn't compare to other words on the page - a trick most of us use.

And, I imagine if doing it 6-8 hours a day, the work done in the final couple of hours was far more prone to error.

I would never trust the transcription unless I have no other choice. In my experience, I haven't had too much trouble with the details you are talking about, but the transcription of names, places and occupations (even where the handwriting is quite clear) is appalling.

[DavidNewton]

I hope that if you do find obvious transcription errors that you will report them to FindMyPast. I have reported a number of transcription errors and it only takes a minute or so to fill in the online form and they will respond. First an automatic acknowledgement and then after a time an email stating how many of the corrections have been accepted.

David

[ADC65]

I hope that if you do find obvious transcription errors that you will report them to FindMyPast.

Yes I do report transcription errors to FMP where feasible to do so, and especially if the transcription is so bad it has stopped me finding someone and I have had to use other search fields, etc., to get to the entry. But annoys me a great deal that I am doing their job for them, instead of paying people properly to do it in the first place. Apart from the privacy restrictions that fhtess65 mentioned, one of the main problems is the images were transcribed mainly by off-shore and quite often non-English speaking transcribers who unfortunately have no context to refer to when reading the entries.

Having worked at data-entry many years ago, I can well remember that we were always told to "type what you see, not what you think", and so some errors are always bound to slip through. But clearly here there has been no validation or double-entry and the amount of mis-transcriptions is off the scale. If the data was free or user-generated (e.g., FreeBMD, etc.) then there is some latitude, but when I'm paying £3.50 for an image that might not be the correct one (and there is no refund mechanism!), it's not good enough.

[jbtapscott]

I would agree with you Adrian although from my perspective the poor transcription efforts from FMP extend beyond the 1921 Census - I have yet find, for example, a complete / correct transcription of the many 1871 Census records I have reviewed with virtually all having minimal "birth place" data and no Occupation reported, as well as some serious Name "translations".
At least with my FMP sub I can freely access the underlying schedule but it does throw up a serious "don't trust FMP transcriptions" view. Perhaps I have been unlucky - most have been Somerset / Devon census records!.

[brianlummis]

If you have not already seen the article, I would recommend that you read "Putting the 1921 Census online - the inside story" in the Lost Cousins newsletter at https://www.lostcousins.com/newsletters ... htm#Rigden which throws a bit of light as to why there are so many errors. The transcribers did not have the benefit of seeing the whole page and the supporting schedules and couldn't always find a similar character on the piece they were using, to act as a comparison.

On the plus side, I have found that any error report I have sent to Find My Past has been corrected very quickly.

[jelv]

Immediately after reading this topic I go to a 1921 census and find this:

1921_Census_of_England_&_Wales_Image_findmypast.co.uk_-_2022-12-04_23.01.02.png (53.73 KiB) Viewed 2433 times

According to the transcription it reads "Both Parents Alive"

[ADC65]

According to the transcription it reads "Both Parents Alive"