Placeholder profiles, good and bad

My intention in starting this discussion is to explore the uses (and misuses) of placeholder profiles in Family Tree. Placeholders are used to gather as "children" the profiles of persons who are not siblings and may not be related at all.
So that no one feels I am pointing accusatory fingers at others, or drawing unwanted attention to them, I will focus on my own work.
I had included images but apparently when a draft is saved any images in it are not saved. Tags are not saved either.
Placeholder profiles in Family Tree
I am working on multiple surname studies. Part of my process is to systematically attach every historical record in FamilySearch to a profile in Family Tree. By necessity, this means I create many profiles. I also fully standardize event dates and places. This makes a significant difference to Find results and hints, especially hints about duplicates. I merge duplicate profiles and split conflated profiles. When splitting I take care to detach sources and delete events that do not belong to the deconflated profile. Because I take so much care with profiles, the research hints Family Tree provides usually are correct. They save me huge amounts of time and effort I would otherwise spend searching.
Merging and splitting tends to build Family Tree significantly, and I get nice messages from contributors thanking me for busting resistant brick walls that have frustrated them. Basically, I approach the wall from the other side. This tree building is exactly why surname studies appeal to me. See this 1 minute video from The Guild of One Name Studies: https://www.youtube.com/watch?v=jFfhh7_OJ_Y
Then, as nearly the last step, I Find and Follow all profiles without a parent or husband of the same surname. (I can find no tool anywhere for this; so this tool is a feature I have requested several times to FamilySearch and also Puzzilla.) I page through Find results, a very tedious process. Years ago I hit my limit of 4000 profiles followed, so I began gathering them under placeholder profiles and watching just the placeholders. I tried placeholders grouped variously by time period, place (nation, state, county, village), and given name. I found given name by far the most useful, so now that is all I use.
I also experimented with how best to name the placeholder. I settled on the given name PROJECT and surname. I put the actual given name in the Suffix field. Gathering by name is most useful because most duplicate profiles have the same given name. If you use the new person page view, you can set an Alert flag on the Note explaining that the profile is a placeholder. See for example PROJECT Crandall Ellis : https://www.familysearch.org/tree/person/details/GXH1-XVZ
A major advantage of using placeholder profiles in Family Tree is the collaboration. All contributors can use them, and I encourage other contributors to do so. If they can find a parent and thus detach a profile from the placeholder, I see that change in my Following list and I thank the contributor.
A major disadvantage of using placeholders is that they create false distant cousin relationships. Anyone is welcome to build any part of the public Family Tree. However, there is a private, confidential extension of Family Tree that is used by LDS members. LDS members are supposed to use that private system to do ordinance work on only their own relatives: their own ancestors and descendants of their ancestors. LDS members are not always very careful about evaluating Family Tree before doing ordinance work. My view is this is entirely on them, not on anyone else.
In the case of household records including a young niece, nephew, or cousin, but the parents of these children are unknown, I simply create a profile for the unknown parent sibling. Often, to do this it is necessary also to create a profile for at least one grandparent. These profiles are not placeholders, merely incomplete. If I can infer an unknown parent's sex and surname, then I have enough information to create a reasonably detailed profile because I can also infer an approximate birth year (I use a decadal year about 30 years earlier than the birth year of the child), birthplace, and/or residence place in the year the child was born. When I make an incomplete profile like this, very often FamilySearch soon offers historical records to document the parent.
Additional companion tools I use to do surname studies in Family Tree
- I use my Following list very heavily. Because I am involved in a lot of different projects, some very large and with many collaborators, I can only follow placeholder profiles, heads of fragment trees not gathered under placeholders, and a select few other profiles with problems such as conflation. If no other contributor does the work needed, eventually I will get to it. Almost all changes I see by others in my Following list are good changes, and when the changes are not good that's a sign the contributor needs support. Either way, my Following list is not a source of aggravation.
- I use the Family Tree partner website Puzzilla.org. A $40/yr paid subscription gives access to a new Research Log tool that is similar in some respects to the Following list on Family Tree.
- To support my collaborations with other contributors I put work-in-progress notes in my pages on the FamilySearch Research Wiki. For example: https://www.familysearch.org/en/wiki/User:Dontiknowyou
- I use Family Groups, although this new feature is so far proving very limited and difficult to use. It seems to be built on top of FamilySearch Messages, which also has issues of usability and reliability.
Comments
-
Well hello! I didn't mean to release this draft to the world yet. Live and learn. Onward...
PROJECT Crandall Ellis : https://www.familysearch.org/tree/person/details/GXH1-XVZ
Here is an example of an Alert flag. The link goes to a Note under Collaborate where the placeholder nature of the profile is explained.
2 -
Here is a screenshot of the descendancy tree view of GXH1-XVZ PROJECT Crandall Ellis. Possibly none of these are duplicates of each other, but perhaps some are duplicates of other Ellis Crandall/Crandell profiles with parents attached. See the hints?
2 -
Well hello! I didn't mean to release this draft to the world yet. Live and learn. Onward...
I think the Vanilla Platform [Save Draft] refresh default may have a hiccup/threshold whereupon it [Post Comment] instead...
Anyway, your thread here is definitely enlightening as to the type of work One Name Studies contributors do/can do in the Tree.
Opinion: Because of
A major disadvantage...
I would like to see FamilySearch formalize a way to enter such placeholders - as kind of 'work in progress' with attached Alert Note. But the current features allow such to be done... I guess with open-edit they are all in progress ... grrr ...
Thanks for the info!
3 -
Family Tree has for the surname Crandall about 30k profiles, of which about 10k still have no parent despite many people working hard on this surname for years. 10k is a lot of tree fragments, also a lot of duplicates.
2 -
I took a run at Ellis Crandall profiles and found parents for all 4 in the screenshots above. But in the process I also found 5 more persons named Ellis Crandall for whom I did not find parents.
At present there are 26 profiles with an Ellis given name and surname Crandall (exact matches): https://www.familysearch.org/search/tree/results?count=100&q.givenName=%2A%20Ellis%20%2A&q.givenName.exact=on&q.surname=Crandall&q.surname.exact=on
2 -
Hah! Knocked out 2 more Ellis Crandall duplicates. They were both Timothy Ellis Crandall KCNT-95L.
0 -
Well, you guys and your merge enthusiasm just inspired me. I added the same alert note to 3 accounts to NOT merge two duplicates unless it can be proven that they are the same boy. I have what appears as duplicate John Micholson/Mikkelsens, but they may not be. One boy is known to have been a son of a set of parents, the other boy is recorded as living with my great great grandmother (a cousin of the set of parents) during the 1880 census and is known to NOT be her son. Astonishingly, I have a photo taken of the mysterious boy and my great grandmother taken around 1880. Word was great grandma never knew who he was. It is plausible they are the same boy who was visiting relatives during the 1880 census, but there is no proof, so merging them would be wrong. A 3rd cousin and I researched the boys until we were stumped and so it remains the best we can do. Both boys researched separately leave a cold trail, so my cousin and I presume the possibility that neither lived to adulthood.
The alerts are in the accounts of the 2 boys and my great great grandmother. I purposely didn't put the IDs because I assumed no one would want to check it over. Let me know if you do!
0 -
My work in Family Tree tends to involve unmerging rather than merging IDs. Like Gail, I have added alert notes to a number of IDs that are constantly being merged, even though I would not consider them to have any strong similarity.
I have not seen duplicates as being the major Family Tree issue - rather it is the problem of separate individuals having been wiped-out due the the carelessness of users who continue to think any John Smith, William Brown, James Young (or whoever) born in England at around the same time just must be a duplicate of someone of a matching name.
I'm sure @dontiknowyou is carrying out important work, but I prefer to spend my time separating, rather than merging IDs.
1 -
Ha ha ha.
I probably spend as much time splitting profiles as merging them. Splitting takes much more time.
Like Gail, I have added alert notes to a number of IDs that are constantly being merged
This is repeating myself, but I cannot say it often enough. When merges keep happening, there is a reason in Family Tree. It isn't other people being stupid (that's stoopid as the Americans say it and stew pit as the English are wont to say it). Tangles beget more tangles. Bad merges beget more bad merges. If a merge is undone but the involved persons, their children, their spouses, their parents, including all these persons' source lists, are not scrubbed clean of residual information that doesn't belong, Family Tree will keep generating bad hints leading to bad merges.
I spend a lot of time on that scrubbing, which might be called going down the rabbit hole, but the result is splits that stay.
users who continue to think any John Smith, William Brown, James Young (or whoever) born in England at around the same time just must be a duplicate of someone of a matching name.
Someone needs to teach these contributors to use placeholder profiles. Don't merge 'em, Group 'em! Or, even better, teach them to attach sources and standardize events. FamilySearch seeded Family Tree with billions of profiles that were barely more than bare names but nobody told most contributors that is not the model we should follow. Nooo, we're expected to finish the job on all those profiles. Who knew?!!! <\mic drop><!-- mic drop is the new soapbox off -->
2