Sunday, May 3, 2009

Vetting Ancestry.com Family Trees

Ancestry.com's interactive family tree database is a nice idea in principle but full of landmines. Researchers make the worst connections, sometimes connecting people born over a hundred years apart and merging children of different families.

I had an email query seeking more information about a listing in my Carr database at Rootsweb for Thomas Story who married Margaret Carr 25 October 1739 in New Jersey.

My information was from this source:
Title: Documents Relating to the Colonial History of the State of New Jersey Vol22
Author: William Nelson
Call Number: F133.N42
This book contains the marriage records of New Jersey.
Bibliographic Information: Nelson, William. Documents Relating to the Colonial History of the State of New Jersey Vol.22. The Press Pringitn and Publishing Co. New Jersey. 1900.

(Links below are for those with Ancestry memberships.)

I really had nothing but a marriage record, so to find more info on Thomas Story, I tried Ancestry's main search engine, which took me to a matching listing in their family tree collection. It showed Thomas Story and Margaret Carr, both born in 1722, with a daughter Elizabeth Story born 1746 died 13 May 1807 in Cranbury, NJ (Middlesex Co). Elizabeth married William Covenhoven, born 2 March 1742 died 9 May 1803 in Monmouth Co NJ. The database lists a bunch of kids, but Elizabeth, Margaret and Anne, born 1764, 1766, and 1770, respectively, seem most promising as their children.

The Ancestry collection is fraught with perils, though, so caution is warranted. The same database that shows Thomas Story born in 1722 says he died in 1873 (at the impressive age of 152) and that he had a second marriage to a Margaret Storie who died in 1873 in Scotland. Also, Elizabeth's children include both Covenhovens and Conovers for some reason. I picked the Covenhoven girls born in the 1760s (listed above) as likely offspring, but there were also children listed who were born in the 1780s, not to mention a bunch of Conover children.

People make odd connections in that database, so pick through it for possible clues and don't accept the data at face value. You'll have to vet your sources. Proceed carefully.