Inconsistency in suggesting "Possible Duplicate Child"?
Margaret MacRitchie ( LDF3-XST ) has a Data Problem (or two depending how you're counting). She and her husband have 2 children named John Constable, both born 1806 but dying 40y apart.
One (1806-1812) is GSLZ-MJN. The other (1806-1852) is LDF3-FZ1. Both appear under her Data Problems, marked "Possible Duplicate Child". And yet I have (just) marked those two profiles as "Not A Match".
Shouldn't the "Possible Duplicate Child" take account of the "Not A Match"? After all, if the two profiles are stated to be different (not a match), they cannot be duplicates?
Yes, there is an issue of having two Johns but whatever that issue is, it isn't a Duplicate Child.
Should the issue be renamed / redefined? is "Clash of Naming" perhaps a better issue name?
Answers
-
Reviewing both records it looks like they have been attached by accident. Click on the edit arrow over the incorrect name and open the edit box where you can remove one of the Johns from the family tree,
I would review the following before taking any action.
Hope it helps…
-1 -
Is it possible that you not only have to mark A as "Not A Match" of B, you also need to mark B as "Not A Match" of A?
Just thinking out loud.
0 -
I believe all you are doing is telling the computer that the hint is not a match; it does not remove the duplicates at all.
-1 -
@ColinCameron said "Is it possible that you not only have to mark A as "Not A Match" of B, you also need to mark B as "Not A Match" of A? …"
An intriguing thought! 😉
However, I can't tell which John I worked from because both Johns have the "Not a Match" in their Latest Changes, so I'm not about to start removing that legitimate transaction and conducting experiments that might mess things up.
2 -
@sc woz - thanks but I personally am certain that the 1806-1852 John is the error and I know how to remove him.
I haven't done so yet because the guy who put him there is convinced that he's under the correct parents. I'm equally convinced that 1806-1812 John is correct because he's mentioned on the family gravestone in Dundee. (And yes, there is a perfectly good reason, with proof, why his parents were in London at that point). Since the Constables aren't that closely related to me and since there are (at least) four other possibilities for 1806-1852 John's parents, I've left him as he is apart from posting a discussion on his profile.
My point in raising this question is not to ask what to do about the Johns but to highlight my belief that the message "Possible Duplicate Child" is inadequate because I've already told the system that the two profiles are not duplicates.
4 -
Thank you, I will see if anyone around has an answer to the specific question.
0 -
Assuming you work for FamilySearch, can you tell us which areas you work in (which engineer teams you liaise with, or whatever) please, so that we know which queries to tag you on (as we do with several of your colleagues) - thank you.
3 -
@MandyShaw1 @sc woz Thank you for your comments and suggestions, but there is no need to "tag" anyone that works for FamilySearch in order to get a query answered. The process would be to post your query or discussion is in the correct Category or Community Group. Then it will be reviewed and answered as appropriate by anyone. Often, other community members (not only persons that work for FamilySearch) answer questions and provide valuable insight and information to the discussions. "Many hands make light work"
0 -
@mynameistk - if you are implying that FamilySearch or their designated moderators review every query or discussion raised in the Community, then that's news to me at least. Indeed, we have been advised things like "For place name algorithms, please tag ...". The current recipient of those tags for that topic has explicitly requested such tags so that she can assemble a list of things to test when the algorithm does get changed. For the avoidance of doubt, I have a great deal of time and respect for someone who assembles such a list. But it does imply that the process of bringing stuff to FamilySearch's attention without tagging is not infallible.
Another point is that my original question has not been answered - presumably @sc woz wasn't able to locate the appropriate person in FamilySearch - that's not a reflection on them, just a statement that it doesn't always work, even with attention from a (presumed) FS person.
1 -
Agree wholeheartedly with Adrian: we've been asked to tag certain people for specific issues. And, there is not a specific category for every conceivable issue that may arise. If there were, each category would have one post only, and then we would need mods or staff members in the hundreds of thousands, instead of the small group that currently handles a myriad of issues.
2 -
I think you have found a flaw in the possible duplicate or the not a match routine that needs improvement. I assume since you have marked the two Johns as not a match that the possible duplicate flag no longer appears on either of them. However, you are seeing the slightly different flag on Margaret of possible duplicate child which suggests that flag does not check for the children being flagged as not a match.
I see, also, that clicking on one of the possible duplicate child flagged does not take you to the merge routine at all. It takes you to the profile page of the child you clicked on where no possible duplicate exists.
I wonder if that is always the case? Or if you would be taken to the merge routine for the two children if they had not been marked not a match?
1
