Better Merger Process
edited February 19, 2021 in Suggest an Idea
David Williams said: The merging process is laborious. Been working on England 1800 records. FS entered each christening with child and both parents. Sound great until that person had 10 kids, and then those kids have a bunch and there are records already made for each, plus a few other people making separate records. No joke. It was 30 merges just on christening records for one family. Then when done, the system discovers all the duplicates because no one wants to merge the many existing records so they create a new person and family with all 10 kids. It would be really helpful to be able to merge child, siblings, and each parent all in one merger. I have spent so many hours merging I forgot what I was doing, just to find there are more people to merge. It has taken so much time and repetitive typing that I want to just move on and forget this family completely.
Tom Huber said: Welcome to the community-powered feedback forum for FamilySearch. FamilySearch personnel read every discussion thread and may or may not respond as their time permits. We all share an active interest in using the resources of this site and as users, we have various levels of knowledge and experience and do our best to help each other with concerns, issues, and/or questions.
Far too many users have merged without checking to make sure the two profiles were for the same person. The process has been refined so that users will take more time in merging two people.
Any the records in the tree were never submitted by FamilySearch. It was users who created those records. They may have created them a long time ago, or in the previous system. Or they may have created them without regard to checking to see if the profile(s) involved already existed within the massive tree.
The current massive tree is the result of pulling all previous systems together into one system. That includes those profiles that were created from extraction projects. They all need to be gathered together through merging so there is only one profile for each person who has ever lived.
Yes that takes time and yes, others had previously gathered the information and created profiles wherein the family had been pulled together.
The current merge process is intended to stop the madness that accompanied users who are in a hurry and quickly merge two profiles that looked like they were for the same person, but were not. The new three step process has proven partially successful, but no system is perfect. Users in a hurry still ignore the differences and go ahead and complete the merge.
With today's process, we now can see more of both profiles, including conclusion reason statements, and make informed decisions about the two profiles.0
Paul said: David
I share your pain - on a regular basis. But, as Tom explains, this problem is down to how records used to be added in the past. I have previously suggested the problem should have been resolved before all those separate IDs (say, ten for the father of ten children, and then more) came across into Family Tree. I now realise this type of argument is far too simplistic.
The past is past and, sadly, we have to "grin and bear" this issue, however exasperating and time consuming it continues to be.0
Adrian Bruce said: Again, I too have huge sympathy for anyone with these families to deal with - it gets magnified when there are not just parish registers to deal with but also Bishop's Transcripts and even Archdeacon's Transcripts... Fortunately, in my experience, BTs and ADTs seem not to have generated profiles very often so they are just a multiple Historical index Record issue.
The problem with trying to combine multiple profiles in one fell swoop is the size of the necessary check. The screens for a single merge have recently increased in both number and their number, in order to push people into really thinking if these are the same. I would defend that increase as a necessary step against "The name's the same" issues.
The problem comes when we try to work out a User Interface that has the increased numbers and steps for everyone. It gets really, really big.
Doing the same "Please think about it" checks for everyone in a family is absolutely necessary. I have several occasions when a husband's second wife had the same given name as his first - pushing them all through quickly could easily result in a merge of the first and second wife. The children might not be a clue, as a census (say) probably won't show that she is a step-mother.
Another potential issue is where names are repeated after children have died but the families to be merged don't have both in - unless you do a careful check, it isn't obvious that Thomas b1879 isn't the same as Thomas b1881, whereas streamlining the merge could very easily result in users thinking that they were.
This doesn't mean that there are no possibilities for improving the family merge process. It's almost certain that the logic for determining that this John and Mary Doe family is the same as that John and Mary Doe family is driven from the match on John, Mary and several, but possibly not all, of their kids.
I'd probably be looking at their individual profiles in separate tabs (I do anyway, even for a simple merge). And then it would be really nice to write just one "reason" for the family as a whole and have that offered again and again for each of the merges of individual profiles. That would, I believe, help us in the sort of case that the original poster describes.0
Tom Huber said: I tend to use a stock reason statement on merges to the effect that "I have examined both profiles and believe they are for the same person." I may add to that to explain why I believe that is so, or I may add, "despite no sources or details in the merge-deleted record."0