Monday, February 28, 2011

FamilySearch begins public beta of "Family Trees"



DearREADERS,
For several years, Ol' Myrt here has recommended that researchers clean up their genealogy databases because it won't be long before FamilySearch will give access to what for years was called "new" FamilySearch, not to be confused with the newly revised website design that we've all been working on since the turn of the new year.

Now comes word from Ancestry Insider that finally, the first hope of public access to the "Family Tree" part of FamilySearch, that here-to-fore only members of the LDS Church have been able to access. AI writes:

"In February 2011, we’ve invited a limited number of public users to begin testing public access to the new FamilySearch [Family Tree] website,” disclosed FamilySearch. “These valued testers will help us make sure the system can handle the increased load.” Don’t bother begging; FamilySearch has already selected the testers." See Ancestry Insiders' NFS Public Release Begins posted this morning.

What's the big deal? 

Well, it will be hard to ignore the elephant in the room when FamilySearch's Family Trees goes live. It will become arguably the largest compiled genealogy database in the world.

This doesn't mean that this database doesn't have it's problems.
My access was limited, since our "temple district" was the last in the world to receive access, owing to the fact that our region of the LDS Church does a lot of temple work. So I cannot report about the progress coders have made with Family Trees.

I can report that there are many problems when combining multiple databases from a variety of sources including the IGI (International Genealogical Index), the AF (Ancestral File), the PRF (Pedigree Resource File), the TIB (Temple Index Bureau) and LDS Church membership records.

In discussions during FamilySearch Bloggers Day in person in October, and via webinar in November, end-user bloggers expressed dismay that the more reliable "extracted" entries were not easily sorted out from "patron" entries in the IGI data. One method for sorting in a previous incarnation of the IGI was to work through the data using Batch IDs, now missing from the new FamilySearch Family Trees.


Why are Batch ID's important? Extracted entries were done by 2 volunteers who indexed original parish records from microfilm, and a third volunteer and/or a computer, (depending on the year), and the work was submitted in batches. So a batch would typically be an entire parish, making it much easier for end-user researchers to find all instances of a surname in the batch, and more readily see index entries that might trace a family through several generations back in time.


Being able to distinguish those 2-3 person evaluated index entries from "patron" entries to the IGI is invaluable to researchers. Problems arise most frequently from patron entries, fraught with errors ro frequently that I personally distrust them completely. (OK, I might look there for clues in a rare attempt at straddling a brick wall.)

Any compiled database from multiple source will have challenges, and the FamilySearch Family Trees is no exception. These are the kinds of errors I see for just one branch on my family tree.
  • Duplicate individuals because of even slight differences in the spelling of names, meant that my great-grandmother's profile in FamilySearch Family Trees has 28 children instead of the 11 we know about.
  • Multiple dates for a single event. You can spot the ones that were calculated from a census record. But in other instances, there are clear typos like 1986 instead of 1886. (Why didn't people run a "possible records problem" report in their genealogy management program before contributing data to the IGI, AF, PRF, and the TI?
  • Incorrect merging of individuals. This happens through indiscriminate use of merging (in one's own genealogy management program) or combining (on FamilySearch's Family Trees) when there are several in close proximity on the family tree with the same given name and surname. Some things just cannot be rushed.
  • Incorrectly associating a document with someone on a family. This is just faulty research and faulty data entry. My Uncle Jack is alive and well. Yet someone had him listed as "female" and deceased. Well, early on, when one could "dispute" rather than "discuss" I objected to the "female" so it was changed to "male" and pointed out that while the Social Security Death Index does have an entry for someone by the same name, about the same age, and a nearby locality, my Uncle Jack is not dead. He just snowbirds to Arizona for half the year.
Last time I checked FamilySearch Family Trees does not allow the incorporation of documents into an ancestor's provile, as we see with Ancestry.com trees. If that had been available when I was working through my Uncle Jack's problematic profile, it would have been easier to prove my point. It would also be easier to evaluate several researcher's contribution to Family Trees if you could view the documents collected by one researcher over another.

FamilySearch Family Trees will not replace individual databases for genealogists. It is merely another collaboration tool. For members of the LDS Church, it can ensure we aren't duplicating temple work on a common ancestor.

It will be interesting to see how this unfolds. As competent genealogists, we must be aware of FamilySearch Family Trees, but take it with a grain of salt. We must also caution the newbies that join our rank not to take too seriously any compiled genealogy on the internet or in a book on the shelf at the library.

All compiled research must be evaluated in the context of cited sources and a reasonbly exhaustive review of surviving record groups for the time and place where the ancestor once lived.

For Further Reading

Happy family tree climbing!
Myrt     :)
DearMYRTLE,
Your friend in genealogy.

7 comments:

  1. Everything Family Search is pushing out is extremely interesting these days.

    Thanks for the in-depth coverage of this, Myrt.

    ReplyDelete
  2. This is a good review. Patience is the key word when it comes to "Family Trees" as it is still a work in progress. Many changes will come in the future as the engineers work to make this "product" more helpful to the genealogical community. In the not too distant future we should be able to make incorrect relationships that can't be broken in the tree as "incorrect" and also edit out those obvious typos, etc.

    ReplyDelete
  3. Excellent article on "Family Trees." Patience is the key word when working with this program as the engineers are working daily to improve the site. In the near future expect to see the ability to mark incorrect relationships as incorrect and edit obviously incorrect information like the 1989 for 1889, etc. Remember the site is a beta site and still developing.

    ReplyDelete
  4. Myrt,Thank you for explaining why's and why-for's to this old gal!

    ReplyDelete
  5. The lack of Batch #s has meant that I have foregone the use of the new FamilySearch, and only use the old IGI. I dread the day when that is eventually taken off line, because there doesn't seem to be any way to sort the wheat from the chaff. Most often I use the Batch #s to search for records within a particular parish. I'm well aware of its limitations, but it's an extremely valuable resource to indicate which parish registers (in the UK) to search next on microfilm (or county Register Office). If this facility is gone, what am I to do?

    ReplyDelete
  6. I may be wrong but isn't the IGI available to search in the regular, recently updated, FamilySearch.org? (By that I mean the newly changed, updated FamilySearch.org, not NFS or the old, since 1999, FamilySearch.org) I thought that the IGI was taken apart and the submitted info put into NFS trees, and the extracted info put into the record search on FamilySearch.org. I know you can't just click on a collection called the "IGI" but isn't that info available in the many different collections like "England Deaths and Burials, 1538-1991"?

    ReplyDelete
  7. You are correct that the IGI is searchable online. However, the ability to SORT or search only by BATCH ID is missing.

    This is problematic if one wishes to deal with one parish in particular.

    ReplyDelete