* Data verification - GRO Index and Certificate

[stripey]

I have created a new fact type - "birth registered" which allows multiple citations to a "GRO Birth index" source and allows recording the GRO index reference data in the "where within source" box. These data are recorded as "secondary evidence" as I will not at this stage have a copy of the actual certificate. The GRO online search provides reliable information and allows me to continue research as at some later stage I will access the actual record and include the image as "citation media" and change the assessment to Primary.
This method works well as the GRO online search is a good tool to verify the birth and death of relatives. However, I have just noticed there is only a partial correlation between the GRO index reference data and the information provided on the certificate which is just the year, district, quarter and mothers maiden name.
Bearing in mind how tenuous some records are to verify is this considered to be a robust method of working?

[tatewise]

The thing to recognise is that the GRO Index and the Certificate are two different source documents.
The GRO Index is just a cross-reference to the page of the Certificate. It is NOT the Certificate reference number.

There is much more information on the Certificate such as parent names, addresses, occupations, etc.
That leads to a whole extra set of Citations for those additional facts.
Therefore, you cannot use the same Citation that links to the GRO Index source.
If you did, then the Citation details & image get replicated in each one, which is not good database practice.

Many users create a new Source record for the Certificate and link its image to the Source record.
The GRO Index cross-reference is usually included in the Publication Info or Title, etc.
A transcript of the Certificate can be created in the Text From Source.
Then create Citations against all the many facts derived from that new Source leaving the GRO Index alone.

BTW: Many users of FH utilise Ancestral Sources to capture those facts and Citations from a Certificate.

[stripey]

Dear Mike, thanks for your insightful response and I apologise for not replying earlier. My reason is that you clarification has caused me to review my whole approach to the issue of certificates.
I first looked at having a source for each certificate (of which there are a number) and as I'm a "lumper' by nature I decided to explore having a single source for all certificates. The reason for this is that in my system the certificates are all scanned and held in computer form. I have therefore established a Birth, Marriage & Death Certificates source which is a computer file which contains scanned copies of birth, marriage and death certificates. The file is located in my Desk Top Computer in file: My Data (D:) Users>Maurice Jones>Family History>Certificates and stores all certificates in .pdf or .jpg format, the GRO Index Reference for each one is also provided. The repository is the GRO.
This allows me to add a citation using this source for a certificate and in the "where within source" field to add reference to this particular folder and individual file (scanned certificate) and to include a media copy if required.
My next question is (as I'm sure others have come up with a similar approach): is this a viable method?
Thanks and best wishes in these troubled times

[tatewise]

To be honest, I have not encountered that approach.
It is very unconventional and is not using the Where within Source field as intended.
You will probably be able to make it work, but will find it difficult to use tools such as Ancestral Sources.

Consider a Marriage Certificate that may contribute a multitude of facts for the Bride, Groom, both Fathers and the Witnesses, including the Marriage event, various Occupation and Residence facts, and maybe Birth events.

Each of those facts will need the same Citation so you will need to create multiple copies of the same details.
If you don't ensure those Citations are all identical then it becomes almost impossible in the future to find all those related facts derived from that one Marriage Certificate.
Your approach is a variant of the Method 2 'lumper' Source technique, but it is not ideal if you need to attach Media or a transcript or notes.
If ever those details need adjusting then each Citation will need to be edited to keep them identical.

What I described earlier is the Method 1 'splitter' Source technique where each Citation simply links to a specific Source record and that record holds the Media, transcript, formal references, notes, etc.
The big advantage is that if ever any of those details need adjusting then only that one Source record needs editing.

[Gowermick]

I can see problems with storing image file Path location in the citation. What if you want to re-organise your files? You’ll then have to modify each citation, a potentially mammoth task. Would it not be simpler to give image a filename (such as the GRO reference) and then just store this reference in the citation.
I do this with census images, so even if I re-arrange my files, I can go to the image’s new location straight away, without having to change the citation at all.

e.g. Citation may read “RG09 Piece 1234 Folio 025 Page 24”, so my file is named “1861 1234 F025 P24.jpg“ and this remains valid no matter how I organise my filing system.

You could use a similar method, with GRO images. So with a birth registered “1886 Q1 Warrington 6c 259”, the image could be named “GRO B 1886 Q1 Warrington 6C 259.jpg”, and stored according to your latest filing system layout.

[tatewise]

So why not just be more conventional and put the Media files inside the Project Media folder using Media records to locate them? That makes naming much easier and if you want to reorganise then there are tools in FH to fix the broken links automatically.

[davidf]

I am also a "lumper" and like the "source" to point to where I got the information - e.g. GRO Certificates, Manchester Council Certificates etc. (they can be different!).

I don't include a pointer to the actual image file in the citation, because that confuses the roles of the source records and media records. The media record holds the pointer to where in your file system you happen to have the image file, and FH supports a limited amount of curating of the files whilst keeping track of them. The citation holds the pointer to "where in the source", (e.g. a link to a specific part of the source record) so for certificates should be the GRO reference (or similar).

Mike always points out the dangers of duplication and consequent (very real) problems of maintenance if you as a lumper should change your understanding of a record. However this is avoidable provided you are clear between the differences between sources and citations.

(For splitters the difference between sources and citation is usually trivial because it is split one citation one source - even though that citation may support multiple facts (marriage date, occupation of father etc.). But they lose the ability to easily hold details of "sources" as you and other lumpers understand them - e.g. GRO Certificates, Beaumont Parish Register, 1881 E&W Census, etc.)

For lumpers where the source is, say, "GRO Certificates", the citation points you to where within the source the particular certificate is found. So like in academic writing where a source might be a particular text, the citation is to a particular page within that text, lumper's citations point to "where within". In that respect the yellow right-handside/bottom of the property box is actually about the citation (apart from the box linking to the actual source being cited).

The temptation is to add media to the fact. But the media does not directly support the fact; it supports the citation. You could for instance have a marriage certificate and a newspaper report with conflicting details; two sources cited with different media supporting those citations. Therefore you attach media to the citation - through the media button in the yellow part of the property box.

It is tempting to transcribe the cited part of the source and put it in the "Text from Source" box. Ideally this should only be used for text specific to the fact. "Father's trade: Joiner" you might insert against the citation of a marriage certificate supporting the father's occupation fact - but you should not put in the full transcription of the certificate. That leads you into the duplication issue that Mike warns you of and consequent need to squirrel around and find all the "text from source" boxes for all the facts supported by citations of a particular certificate. (Suppose you later decided it did not say "joiner" but "jailer"?).

So where to put the transcription of the certificate, if you only want to enter and store it once? A shared note is the answer. Unfortunately to attach a shared note to a citation is a touch tricky in FH, but can be done through the "All" tab on the property box.

• Within the property box find the fact being supported by the citation (remember a marriage fact is not an individual fact but a family fact, so you have to click through to the relevant "family as spouse" and then find the marriage fact)
• Work down the tree (widening the record column if necessary to read the branch names) to the source being cited (i.e. the citation!)
• right click on the word "Source" and select "Add Note", "Add link to New Note Record"
• The full transcription can then go in this note (just give it a useful title either in the Custom ID field or in the first line)

Marriage of Joseph Sewell Faulder to Emily Story - Citation & Media

Marriage of Jospeh Sewell Faulder to Emily Story - Citation and Media.png (57.15 KiB) Viewed 4496 times

(In the above you will note that I have "Certificates: Marriage" as my source - when originally entered I had not twigged that a GRO certificate could be different from a copy from the actual register.)

Note that the above shared note is not a source note but a note attached to the citation. Lumper's still have the option to use source notes to hold comments specific to the source (which might be about how to obtain certificates, your current strategy for when you bother to get a certificate, or specific details about say a particular council's registration service "ask for Ian, he knows his way around the system"). If splitters want to hold such information, they either have to duplicate it against every split source (with the same maintenance issues) or they have to use a shared note - which they can then attach to multiple split sources. You takes y'r pick!

You then have a citation (and media link) which you can copy and paste (using the - correctly named - buttons in the yellow part of the property box), but you only have a single copy of the transcription held within FH. Duplication and maintenance problem solved.

[stripey]

Many thanks to Mike and to everyone for their contribution.
After consideration I consider my best course is to follow Mike' advice as I have no wish to build in a future maintenance problem. So a lumper I may be but pragmatism has driven me to a hybrid approach.
Best wishes and lets hope we can soon return to normal.
Stripey

[tatewise]

Excellent. The point to recognise is that users should NOT be exclusively 'splitters' or 'lumpers'.
That choice depends on the Source Citation characteristics.
GRO BMD Index records lend themselves to being Method 2 'lumper' Sources.
BMD Certificates are usually better suited to being Method 1 'splitter' Sources as they often have associated Media images, Text From Source transcripts, and multiple Citations.
So most users adopt a hybrid approach as you have done.

Popular criteria for choosing between Method 1 and Method 2 is in the how_to:key_features_for_newcomers|> Key Features for Newcomers cunningly entitled Sources Methods 1 & 2.