Has anyone written up their "best practices" methodology of searching for and merging people with a
I have an ancestor without known parents, and I am trying to "hook them up" to generations further back in Family Tree. I search for their surname, and find many "leaves" ... FS ID persons of a pair of parents and one child, no sources attached. I think it would be neat to correctly piece together these many many leaves into branches of the tree, and perhaps I can then find the linkage to my ancestor's parents. Here is where it gets "interesting" ... by which I really mean "daunting."
My surname of interest has about 8 main spelling variants. Place names are 50/50 either in Slovakia or Hungary. I've used the Slovakia Genealogy Gazetteer, which is wonderful. I have tried FIND commands with Exact Match and without. Sometimes, Exact Match is not supported for certain data fields (such as birth year ranges). FIND returns either too few records or it returns thousands. I have tried exporting Search command results as XLS format, but there is no linkage between that XLS data and Family Tree persons. So, while I have now done some Merge activity and attached Sources activity, I am at a loss how to optimize my efforts, and the task I seem to have in front of me looks like it will take many months or years. Can someone describe the way(s) that they have found to be most efficient ... at a summary level ... to make progress like this?
Answers
-
I think during you build your family tree it will be "best practices" and you all right " it will take many months or years".
But if you have a specific problem we will try to help.
Otherwise "Search Historical Records" sometimes not easy.
My main problems are similar spelling variants or misindexed records.
0 -
Those little family tree twiglets you've been finding are based on some of the indexes that you can search under Records. Not all index entries have such associated twiglets, but all such twiglets have an associated indexed record. (In my experience, girls with indexed baptisms almost always have Family Tree profiles, while boys very seldom do.)
In order to deal with the profiles effectively, you have to understand the nature of the profiles and the records they are based on. This is legacy data: it was transferred from FS FT's predecessors. Those preceding systems/databases did not implement the current "one tree" model. There was no attempt at combining the records for parents who had multiple children in the system.
The records forming the source for these profiles are indexes of church registers. The indexes were made by amateur volunteers. The volunteers often had no familiarity with the language of the records, they had no special training in the interpretation of handwriting, and they knew little or nothing about the sorts of names to expect to see. Therefore, they made mistakes. However, the process by which the indexed records were turned into profiles ignored the possibility of error: the profiles use the indexed names exactly, letter for letter.
When I encounter a twiglet of interest, my first step is to attach the source. FS has often helpfully already found it and offered it up as a record hint, along with record hints for the girl's probable brothers and possible duplicates of the parents (each with their own record hints). I do not attach the records for the possible brothers yet, but I open the possible duplicates in new tabs and attach their source records.
Once each twiglet has been linked up with its originating index entries, I start looking at the record images, and I try to determine whether the possible duplicates are in fact duplicates or not. I check things like the father's profession or status, the house number, and the names of the godparents. If these are all the same, then I can fairly confidently conclude that the baptisms are for sisters, and I can merge their parents.
Things are seldom so easy, of course, so eventually I end up writing up lists or charts containing all the children that Search - Records is willing to cough up with the same parents in the same place. I suppose I could use the export function to start this process, but I don't: the indexes don't contain any of the cross-check data that's the whole point of the exercise, so I find it easier to just start from scratch.
Because there are so many errors in the indexing, coming up with a full list of children can take some creativity. Sometimes I've resorted to just browsing the images; that's how I found Alojzia Gröbersperger, who is indexed with her father's name ignored, her mother turned into her father, and her father's occupation and religion turned into her mother (https://www.familysearch.org/ark:/61903/1:1:QVNX-3382).
The "too long, didn't read" version: start on the Family Tree end instead of the Search - Records end, and use the hinting system to get a start on assembling each family.
0 -
Thank you very much, Tamas and Julia. I am grateful to read your advice and experience. Your answers tell me one important thing ... this is clearly why coffee was invented ... to help keep genealogy researchers awake for months at a time! Thank you again.
0
