Tuesday, February 13, 2007

Matching marriage entries, and other news

I have almost finished sorting out the Liverpool district Parry marriages, to help Susan with her challenge. The process is not really complicated, just time consuming, as I have mentioned before. Basically it involves:
- already having a listing of civil registration items (perhaps extracted from something like Freebmd, or transcribed personally from the GRO indexes)
- collecting the items on the Lancsbmd site (relatively easy because it is possible to download a csv file of the results).
- checking each item from Lancsbmd to find its GRO reference (using Freebmd as far as possible but looking up on the original images where necessary). [For a low frequency name, it might be possible to skip this step and to match to a listing of the civil registration items just on the basis of the names alone but this is not possible for the Parrys – they’re just too common.]
- matching the Lancsbmd listing with its added references to the listing of civil registration items.

In some ways, it would be easiest to run the match process using a program like Access but there are some difficulties with doing this. Firstly, if running the match on the basis of just the GRO references, then there are additional entries created in any situation where two Parrys have the same references (which can happen because of Parry-Parry marriages but there is also at least one case of two Parrys marrying non-Parrys yet having the same references. Was it a brother and sister marrying on the same day or was it just coincidence that they followed each other in the register?) Although it is possible to automatically delete any duplicates, this won’t solve the problem of "false matches".

But, if the original matching process takes names into account as well as references, then it won’t match items where there are slight differences in spellings, such as the Susannah who appears on Lancsbmd as Sussanah, or the James who probably should be a Jane. And it certainly won’t match the "Christina" who probably should be a "David John"!

Although the Freebmd name transcription can be checked back to the original indexes, there’s no way of checking the Lancsbmd entries. And what if the transcription is correct but the spelling is just different in each source? Such situations lead to the question of whether I really need to be adding a standardised name column to every file as well (which I will obviously need eventually, when I finally want to put all of the names into a master index. But do I want to sidetrack onto consideration of "standard forms" at the moment?)

And, of course, any automatic matching process using the GRO references, presupposes that the references on the civil registration listing have all been corrected (which they haven’t!)

In the end I just find it easier to match using Excel, keeping all the columns of each file but then also having one column with all the GRO references in to sort on, which brings the matching items next to each other enabling them to be combined manually. It might take longer but I keep control and at least I know there are no false duplications. Then it is just a matter of putting the information into a form that is usable by Susan since I have sent both a full column file (in case she wants to double check anything) but also a reduced file based just on the references and including spouses and church, which should be easier for her to work with for the challenge.

Hopefully, having now listed all the process, it will remind me of what to do if much time lapses before I tackle any more UKbmd sites.

Other news - more years have been added to the passenger lists on findmypast – it now covers from 1890 – 1909 and includes 2038 Parrys.

Another researcher mentioned a strange entry on FamilySearch, where what is possibly her Williams family have all been entered as if a Caleb Parry is their father (easy to find – use "All resources, look for a Martha JASPER, with Caleb in the father’s name. First entry, on the IGI, for Martha Williams Jasper, born 14 FEB 1821 Llanbedr, Brecon, Wales, is the relevant one. Click on her and then the "Family" link alongside Caleb.). Having done some further searching, I found an entry in the 1871 census for Abergavenny showing a Caleb Parry visiting the Jaspers. Caleb was originally recorded as a Minister, to which someone has added "Latter Day Saints". Now I wonder when that was added!

Caleb is recorded as being born in America but, based on various census and IGI entries, I suspect he might actually have been born in Flintshire, and that he then moved out to Utah, marrying a Catherine from Wales either before or after he arrived there (or on the boat, as some IGI entries have it). He then returned to the UK around 1870/71, and died in Birmingham in 1871. There’s also an interesting collection of other marriages, along with some children born to one of the other wives, children who do appear with Caleb and Catherine in the US census. I wouldn’t have been able to tell from just the census entries but it seems possible that this is my first case of the "multiple marriages" one is reputed to find amongst the LDS.

Funnily enough, one of the names which appears in the family is Bernard, and I have just dealt with a query from another researcher who has family using that name, in the same area of Flintshire. I wonder if there might be a connection there? It’s a pity there are so many of these North Wales families – it will be a fascinating job to sort them all out eventually but, at the moment, I just don’t have the time or resources.

Next priority is to update my website, and add that link to Ruth's site.

No comments: