Saturday, July 19, 2008

Excellent CGSSD meeting today with Joel Weintraub, Ph.D.

The Computer Genealogy Society Of San Diego (CGSSD) meeting today commemorating the 20th anniversary of the founding of CGSSD was one of the best I've ever attended anywhere.

Due to another commitment, I missed the first 30 minutes of the first hour of the program. I was told that CGSSD President Corlee Morris introduced the Charter Members in attendance (there were 30, and some are still active), the Past Presidents (many were there!) and honored guests (they invited the Presidents of other local societies - a nice touch). There were presentations about CGSSD history by Corlee Morris, computer genealogy history by Dona Ritchie and on the future of computer genealogy by Gary Hoffman.

Joel Weintraub, Ph.D. was the featured speaker today, and he and his presentation were geneatastic - (defined by me as a genealogy event rated somewhere beyond 11 on a scale from 1 to 10). There are really no words to explain how good it was. Joel is knowledgeable, funny, an excellent communicator and he put together an informative and entertaining PowerPoint presentation. The audience experienced his many years as a college professor and a genealogy researcher in his presentation.

Joel's topic was "Searching the U.S. Census by Geographic Means When Name Indexes Fail." He provided a brief history of the US federal census records. He passed around some examples of actual census pages, enumerator applications and handbooks. He provided some "best-estimate" statistics for the Undercount in the 1880 to 1990 census records, and some reasons for the under-count. One of the Three Stooges, Larry Fine (actually Louis Feinberg) was a case study about the 1930 census and Ancestry indexing methods. Joel talked about the 1940 US census, and why it is the beginning of the end of "genealogy census" records - due to fewer questions asked of every person and more asked of a sampling - in 1940 the sampling was 2 lines out of 40 on every page - 5%. Lastly, Joel discussed his research at NARA-Laguna Niguel in the census records, and why and how the Steve Morse One-Step pages (at http://www.stevemorse.org) were set up and how to use them effectively to find Enumeration Districts in the 1880 to 1940 census records. Joel provided a three-page handout for the audience.

At the beginning of his talk, he asked the audience "how many of you do mostly "exact searches" on Ancestry.com, and how many of you do mostly "ranked searches?" He was surprised that the response was about 50% each (I'm not sure what he expected) - I raised my hand that I do mostly "exact searches." He has a slide in the presentation handout providing search strategies for exact and ranked searches. Even though the topic was "geographic searches" Joel didn't spend much time discussing them - he had a slide in the handout.

For me, the most interesting information imparted were:

* The undercount rates for the 1880 to 1990 census records. The undercount was in the 6.5 to 7.5% range for 1880 to 1920, and about 5% in 1930 and 1940. He said the 1870 census was the worst because of fraud and semi-literate enumerators, and that many regions in the South were 10% undercounted.

* Ancestry.com weights the last name over other search parameters in their non-exact searches.

* Ancestry.com changed their "ranked searches" to a "sorted by relevance search" recently and the search results are different. In his case study of the 1930 census, "Larry Fine" wasn't even on the list of a ranked search before June 2008, but is #1 on the "sorted by relevance" of a non-exact search. Why? Because he was initially indexed as "Lany Pine." Joel initially found him by using Fine's wife's given name, Mabel, and Larry's age, birthplace, and residence (but not Larry's given or last name). At some time, someone added the name "Larry Fine" as a correction to the Ancestry index.

* For the 1940 US census, NARA will not microfilm the population schedules, they will scan and digitize them; he said that NARA is looking to a third-party provider to scan and index them.

Joel has developed a web site with links to census records, Ancestry databases, the Steve Morse One-Step site searches, and more at http://members.cox.net/census1940/.

This was the best genealogy presentation I've ever seen - either in-person or online. Joel's communication skills - both verbally, technically and graphically are excellent. If RootsTelevision folks are reading this, I suggest that you video Joel's talk(s) if he will allow you to.

Needless to say, I highly recommend that you attend any presentation given by Dr. Joel Weintraub - even if you think that you are an "expert" in the subject.

No comments: