Merging
So, how many of you are tired of people merging your ancestor with a completely different person? Ignoring all of your attached sources, life sketch, notes, memories, etc.
I have recently started contacting some of the people that have done bad merges, asking what prompted the merge and explaining what they missed that should have tipped them off they were not the same people.
I find most are beginners, a couple didn't realize it wasn't "their" tree and affected everyone else. A couple were just nuts..."Well fur traders had multiple wives" This person merged every James McKnight he could find with the name Mary together. It did not matter that some were in Ireland, some England, some California, some PA..get my drift. He went on to tell me how "helpful" he was being. I explained he had merged over 10 families together and it took me over 2 days to fix and I probably missed some.
Another merged a bunch of families together using just the father's first name. There were 4 or five different surnames involved, I had to contact 2 others on FamilySearch to help me correct and again it took over 2 days and I am sure I missed a lot on that one. But I did my best to separate them at least back into their appropriate family units.
I HAVE AN IDEA, maybe FamilySearch should require everyone to do a tutorial on merging prior to giving access to this feature. It could be a simple online tutorial. They already have the How To section explaining the proper way to merge and why it is a complex process that should be done with care. It could be as simple as requiring them to read that.
I am sure there will still be errors in merging, but, maybe it would cut down significantly the number of bad merges if beginners did not have access to this feature without first learning a bit about the difference between your "own" tree on sites like Ancestry and a collaborative tree. And why it is so vital to read through the Source, Collaborate and Memory section PRIOR to doing a merge. And then paying close attention to EVERY SECTION on the merge page PRIOR to completing a merge.
I for one am very tired of wasting hours restoring people, moving children around, detaching and reattaching records, etc.
Thank you for you consideration. Any other ideas?
Comments
-
PS To those of you that have done a lot of work on your tree but may not have attached all available sources to your people. I have found that attaching all available records to my Direct ancestors and adding detailed life stories in the life sketch has greatly reduced the bad merges on my direct ancestors. I still run into problems with some born prior to 1700 (seems to always be someone that found a family tree on some random website without sources and assumes it must be correct). I recommend adding as much detail as possible in the detail section and life sketch rather than the notes on any ancestors that you repeatedly have to unmerge.
4 -
Several other users have offered similar advice - in particular the request for some sort of "mandatory training" before a new user has the ability to make merges.
However, as the years past (I believe Family Tree has been around for around ten years now) it seems highly doubtful that this suggestion will be implemented by FamilySearch.
In respect of adding as much back-up to each record (ID) as possible, I agree this is really important. But I have found little evidence that it will deter many users (inexperienced or otherwise) from making the most absurd merges you could imagine. I, too, spend up to 2 or 3 days from time to time in undoing the damaging results of multiple merges - often involving even a different name, let alone the different individuals being from completely different parts of the planet, and even from different time periods!
Again, as you have probably found, responses vary from none at all, to a simple one word "Sorry!" I gave one person the benefit of the doubt in politely suggesting she might be inexperienced in undertaking this type of work, but she responded that - no - she had been carrying out merges for many years!
Most of us "everyday" users (literally, "everyday" in many cases) have grown quite used to the inconvenience of carrying out lots of "remedial" work, but accept the continuing benefits of using the website outweigh the negative aspects, so despite having our regular moans we then just carry on, in the knowledge there will probably be little change in the situation.
6 -
I'm leery of any suggestion that'll introduce yet more obstacles into the merging process, because I deal primarily with the other side of the coin: what I refer to as "index-based auto-duplicates". These are the legacy profiles imported into Tree a decade ago, where if a couple had eight children baptized, then there are eight copies of each of them -- or more, if the page was filmed twice and therefore indexed twice. (Differently each time, of course.) Once you get to kid 5 or 6, the process starts yelling about too many merges, and the whole thing is just tedious beyond belief, especially when there are duplicates of the parents based on their baptisms, plus duplicates of each kid from the grandkids' baptisms, plus duplicates of the spouses, and so on and so forth, ad nauseum. It doesn't help that none of the canned reasons even remotely works for these types of merges, or that there's no good tool on FS for recording the addresses and godparents and other details from the images (to avoid merging similar-named families).
I could perhaps get on board with a per-account "merging permission" setting dependent on the completion of a one-time merging tutorial and quiz. For the quiz, I envision something like three carefully-constructed (but realistic) comparisons: "based on the information seen here, should these profiles be merged?" One of the comparisons could have identical names but different continents, one of them could have identical names and locations but different centuries, and one of them could be small spelling variations of the same names, places, and dates.
Unfortunately, I highly doubt that FS will go this route, or anything like it, because the tutorial and quiz would need to be translated to every interface language, and the accounts-and-logins setup would need an extra tier added. (We'd need public, public with merge, LDS, LDS with merge, plus whatever administrative tiers already exist.)
6 -
Amen to that. Some random person finds the name of your ancestor married to somebody and, without thinking or researching it further, they add the new someone as a spouse.
This happened with one of my great-great- grandfathers. Someone found his name (which is a fairly common name for that time) in a marriage record to a woman in England, so they just popped her in as his wife. Never mind that he never married in England -- came to the US as a young man. The person who did the adding was totally unrelated to him -- just found a record. So I unlinked her as a spouse and added a note not to add any other spouses than what was in there right now.
One of the strengths of FS is that anyone can make a change to anyone. One of the weaknesses of FS is that anyone can make a change to anyone. I've found that I have to periodically scan through my recent ancestors / relatives and look for changes, and undo the crazy ones -- because there are a lot of careless, unrelated people who are not conscientious about what they do. That's just a hazard of Family Search, Don't know how to get around that and still have the ability to accept valid input from a lot of people.
2 -
The best thing WE can do to avoid people playing around with the work is to: source it to death. FS seems to not want to discuss what the primary users want to discuss about a problem. The feeling seems to be, at least in most circles: ALL revelation comes to the head and the head only. I like to think it comes sometimes from the everyday peon.
1 -
I just starting using wikitree. They have a Pre-Cert process for anyone that wants to edit people born prior to 1750. It involves first being on the site for more than one week, next, making at least 50 contributions, and a short quiz that you must pass.
Prior to achieving that status your ability to edit is limited.
Seems like something FamilySearch may want to consider. At the very least weeding out newbies.
1 -
Paul W Re: "making the most absurd merges you could imagine."
I was trying to be kind and give people the benefit of the doubt. I have run into some inexplicable merges and edits to sourced data. Most recently someone's ancestor was linked to the wrong family line. Instead of removing that one person, the user changed the names and vitals of the parents, grandparents, great grandparents, and etc. to "make it match" their Ancestry tree. Ignoring an abundance of sources attached to each and every individual. So now their one relative had the correct ancestry BUT every other member of that family (they had several children) now had an incorrect tree. To add insult to injury, the family that their 1 ancestor belonged to is an existing family already on FamilySearch, also very well sourced. All they had to do was move that one person to the correct tree. I spent hours fixing that, first I had to try to decipher just what they were trying to do and why (because of course they sourced nothing and the reason given for all changes was just "Warren is on wrong line".
1 -
Julia Szent-Györgyi wikitree had pre certification quizzes plus other things one must achieve prior to being given to access to edit pre-1700 and pre-1500 persons. I think it is something FamilySearch should really consider and include a pre certification quiz to merging feature.
The damage to the accuracy of family lines is seriously compromised by unconscientious individuals that merge without regard to the affect it has on multiple families and without regard to existing sources. I often wonder if the individuals doing these sort of merging really understand that they are working on one collaborative tree.
1 -
I am watching nearly 4000 heads of tree fragments, meaning profiles with no parents attached. So they are ready candidates for bad merge hints. Yet I see a bad merge maybe once every two years. Why are bad merges not a problem for me? I think it is because I fully standardize event dates and place names. I do not use the half way "shadow standard" practice so often recommended here.
0 -
@Melissa Ann McLaughlin, no, the sort of person who hijacks a dozen profiles in the name of "Warren is on wrong line" clearly does not understand that it's one collaborative tree. He's doing what he'd do on Ancestry or MyHeritage if he found a mistake.
@dontiknowyou, I highly doubt that using only verbatim copies of standards database entries has anything whatsoever to do with a lack of bad merges. The hinting algorithms only look at the standardized versions, and clueless users only look at what the hinting algorithms tell them.
1 -
The hinting algorithms only look at the standardized versions
That may be, but when the standard is only suggested, not accepted, the matching is very relaxed and the result is bad hints.
0 -
Late to this, but @Melissa Ann McLaughlin's experience mirrors many that I've had. I mentioned in another thread how often I've run across people who mass-merged absolutely everybody with similar names in a certain area -- meaning, for example, having "John" as a first name and a surname that starts with Mc in Scotland (or Canada) sometime in the 18th Century, then doing the same with all their wives, whose surnames were often unknown. These are at the very least negligent, and more likely intentional. Nobody paying even a modicum of attention could miss the conflicts, or sincerely believe that one man married 30+ wives and had 80+ children.
Yes, these editors frequently think they're working on a private tree, but that seems like an even better reason to block editing access for people who have made irresponsible mass-merges, at least until they clearly understand how the tree works and acknowledge that, yes, there might actually be more than one person named John Wright in the 18th Century.
1 -
In the old old olden days people had to run their work past a ward auditor (not sure of title) before submitting names etc. Melissa, don't completely give up on the certification idea. I really like it...especially if they are already doing something like it in Wiki!
0 -
Thanks everyone for your feedback. I work on this sited daily. I run across bad merges almost daily. The minor ones are easy enough to fix. I have been working a surname project in Phila. for 3 years. It is extremely frustating because when I work a family I attach all available sources and do research beyond FamilySearch and add those sources (transcribing the details of my source). To then come across a family I had correct to find that it is now a combination of 4 or more families. It takes hours or days to fix because after so many merges and so many attached sources I then have to go back and read every source all over again to figure out which kids go with which couple. Yes, I restore the deleted record, but then you still have to go back in and detach from the incorrectly merged profile. PLEASE FamilySearch, do something about this problem. It is beyond frustrating to spend weeks researching a family to have some come along and obliterate it.
In the worst cases the person not only merges everyone, but they change names and dates to "make it fit" their unresearched online family tree they are copying from (usually Ancestry). I know this because I have been able to track several uses back to their unsources online Ancestry tree (or oversourced Ancestry Tree). When I say oversourced, those are the people that attach every source that pops up as a hint without actually reading the source. These people may have 2-3 death records attached to one person. I welcome newbies and beginners. But, when I was a newbie I read every genealogy book I could get my hands on BEFORE I started working online.
2
