Conflicting Data- Comment from a Curator

Marv -

Thank you for sharing this.

(It would be a great comment/discussion for the Collaboration Pool Project... Do you mind re-posting it there -- or, may I?)

Go a head and re-post it.

For what it's worth, here are my stats:

340 Collaborators, 31,822 Tree Conflicts, 38,961 Data Conflicts
Resulting average of over 94 Tree Conflicts and 115 Data Conflicts per person!

I usually work from the filter views of Managed or Curated by Me. I was just beginning to see a light at the end of the tunnel.

Switching to view Managed by My Collaborators was... like a cave-in in that tunnel. Time to dig out!

Conflicts are always the problem...big or small.
We are dealing with dates, spelling of names, generations, children, wives,
husbands, places born, died, baptized , in-laws, half-siblings, multiple spouses, etc.

That creates countless opportunities for mistakes to creep in. In addition, we often do not have complete information because damage to records has erased many records that might have clarified the answers to problems.

And then, even though we have access to millions of records.....we do not have access to ALL records. In addition, much information was never recorded and so is lost in time.

So, it really is not surprising that we have so many bloopers and mistakes
to correct. Sometimes the mistake may be simply a typing error. Again,
it may be that the mind gets tired and flips images in the brain.

Just ventilating and pondering on the magnitude of trying to make a perfect tree. Probably such a dream does not exist.... but we still keep trying to repair and correct wherever we can.

Private

I understand that a perfect tree of 52 million people is just not possible.

I opened up the first 50 tree conflit in my list, and that was all it took to fix 90% of them. The computer just needed to set the parents. I didn't have to do anything but open the tree on that page.

The Data conflicts on the other hand is another story, did he spell his name brown, browne or braun, Caulk, Calk or Cawlk. And when you hit native americans you have an English Name an Indian name a translation of the Indian Name, the spelling of the name using the Cherokee alphabet, was he a Chief or a Uku and.........

All i'm saying is if you have a data conflict and 8 managers, try to get all 8 to agree on a common name and put the rest in "Nicknames" and the "about me" and "discussions about this person" can be a reccord for why the persons name is the way it is.

The tree is ever changing/growing. The system is set up so you can show other people how you came to the way the tree is with sourses, discusions, videos and documentation. If the only thing you have is "my father said when I was small", document it. It may be the only info on Uncle Ed. Those papers you have in your Garage, copy them and upload them to the correct profiles, they may be the only record of a person.

We all have that ancestor and that brick wall. If we all put that little bit of info and put it together we can do a lot better job.

Thanks for listening to me rant on. It is not that I think that people are ot doing the best they can. Many people do not even know that there is conflcting data in there trees because they don't know where to look. They may not know that there is a tree conflict and wonder what that yellow triangle means in thier tree. There are people that can go thru generation after generation correcting the tree as they go and others need help just to get thier children in the tree. This discusion was to help educate not to make someone feel hat he/she is not doing the best he/she can.

Please vote/coment on this
http://help.geni.com/entries/509316-correcting-data-conflicts-when-...

I truly believe that geni is presenting all of us with a wonderful and unique window into history - our own families and the larger picture. To be honest other genealogy programs look childlike and simplistic to me after working with geni.

That said, it's a robust, powerful, subtle and ever evolving application, with upgrades on the "fast track" and data increasing exponentially moment by moment (just watch the "World Family Tree" screen for a few minutes and you'll see what I mean.)

That also means that every day new users try and figure out their way around, and that's a learning curve. I'm pretty good learning applications and I'm still discovering more nuances and "best practices" a year later.

So I for one appreciated Marv's reminder and went through my "data conflicts" screen. Have it down to 30 I think!

BUT no WAY Im checking "with collabs." I'd never dig out.

Marv

Voted and commented. To be honest I see it both ways and the ideal case is to be prompted and make a decision -- progress with merging or go resolve data.

Dear Marv,
Thanks for expanding on the problems of merging and correcting conflicts in various family trees. I am improving in my efforts to merge various profiles by comparing the data. Many people are also taking a careful look at the differences and trying to help resolve those differences. As you say, progress is being made by many collaborators and curators.

Whenever I find a conflict that disagrees with the information in my genealogy program, I try to go back to the internet and seek the best
sources for a resolution of these discrepancies. Some websites give many sources for their data.

I have found a goodly number of mistakes in my own records as I scanned through the work of others. And so, I correct my own history as
I am trying to match up information on merges.

I agree with you and Erica and Jenna that this Geni project is an amazing piece of modern wizardry. The people who program the Geni
have tried to create helpful tools for managing profiles and keeping the
tree as accurate as possible. The Master Profiles have become a great
step forward . In the process of checking the accuracy of the Master Profile, a great deal of questionable conflicts can be resolved. I still find
many errors in the number of children who are attached to these Master
profiles. However, overall, these Master profiles have provided the researchers with the best, most dependable data that can be found.

Of course, we cannot find original sources for every piece of information in a profile. Researchers really try to study the records that have survived and then judge the validity of that material.

The histories of our families are not guaranteed to be 100% true and
trustworthy. I am convinced that most people who trace their family history have learned that it is difficult to get a complete and accurate
picture of all of our ancestors. It is remarkable that such a great mass of
records has survived such as the censuses and the birth, marriage and death records of the churches, villages, and cities.

I expect that we, as compilers of family data, will always be pleased by the goodly amount of information we find......and also disappointed by the holes in our history where information has simply disappeared.

IMHO, just checking on the green may be an initial help, but using some common sense when having the 2 similar profiles on display prior to requesting a merger is probably the best idea.

First of all - if any - check the dates of birth and/or death. They may not always be accurate and while the death dates tend to be more precise (legacy profiles), the birth dates - if any - are often but pure guesses.

The living status is often wrong thus indicating that some of the profiles are several hundred years old...

Next, try to compare the names, often long with embedded titles or fifehoods, sometimes in different language versions...
Also the parents, for some reason more often the mothers of the profiles give a helpful hint, so compare those.

But...be aware of the possibility, that sometimes several siblings of same gender carry exactly the same names while still being different persons: While we seem to shy away from this in modern days, it was not uncommon to give a newly born son exactly the same name of that little one, who just recently passed away... Again, with dates, birth or death, attached to the profile, recognition is easy.
A very common and in a way similar problem is the use of the same names through several generations, being a little bit sloppy here could result in a merger of father and son - or -worse - and that tends to have serious and difficult to repair negative consequences on the tree(s).

Going to the sources always is the best solution, but indeed a very time consuming one :)

Just a few ideas,

/Sten

They are excellent hints Sten! I also love Ethel's point - researching family history may always be a work in progress, including "brick walls" and having to accept the reality that there *is* no record, alas.

One thing I really find helpful is a link within the "about me." That way I can call up another screen and compare.

@ Sally Hines Please tell me how to post half siblings?

Sally this sounds like an incomplete merge. In tree view do you have a yellow triangle? Click on it, the program will guide from there.

If that doesn't help post a link to the profile or start a public discussion from it & someone will help.

Checked the Collaborators section in the Activity tab of your own profile?

If you click View All you have the option to search for and remove collaboration.

http://www.geni.com/list/collaborators

Your own profile > activity tab > collaborators > View all. That pulls up a list with a "search name" box.

I looked under Family > lists and didn't see it there. But of course then safari on iPad crashed.:)

Yes, me either. I had to ask Bjørn (and he's reminded me before!)

What is the logic behind conflict resolution approval?
Sometimes i can Save the resolved conflict, other times i have to Submit it to the manager (who potentially hasn't been active on Geni for years).
What factor determines if the Save or Sumbit button appears?

Probably "collaborating" permissions Alex but I need to "ask Bjørn" to be sure. Also the data conflicts should appear in all the managers merge centers, as well as showing, in profile view, on the menu Actions > Resolve conflicting data.

That was a good question. As a curator it appears I can resolve any data conflicts.

I have never seen a "Save the resolved conflict" option in the conflict resolve window.