30 October 2012

Those Suggested Records at Ancestry.com

The theory behind the "Suggested Records" at Ancestry.com is a good one. However, like many good theories, practice is something else.

The 1900 census entry shown is for James Rampley and his wife Nancy E. James had a brother Riley. The Rampley brothers, and their families, are fairly well-documented and, not all that difficult to trace (except for the fact that they both married women named Nancy Newman).

However, a researcher in their tree has "merged" some information on Riley and James. Because the "Suggested Records" pulls linked items from the trees, the suggestions for James now includes items that are Riley and are clearly Riley and don't include a James born about 1835 in the entry.

I sometimes have used the suggestions to give me ideas. The more that these suggestions are pulled from user trees, the less likely I am to use them. It would be one thing if the suggestions were only pulled from records that appeared to have a "close" match to the desired person. But if the suggestions are pulling in items that users have linked to their trees, then the mess has become even bigger.

No comments: