Barking Up the Right Tree: Shaking a Few Good Hints Out of Ancestry’s Shaky Leaves
by Brenna Corbit, Technical Services Librarian
In the last installment, I went on a rant concerning the dubious workings and shortcomings of Ancestry’s shaky leaf. But this time I will show you how to shake a few relevant matches from that miscreant leaf.
 |
In the tree view (top left), hints appear as a shaky leaf.
But in this partial screen shot example
(bottom)
of the profile view, the shaky leaf hint appears
as a green spot. There are two hints available for this
individual.
|
I have pointed out in previous articles the issues of indexing, exact versus fuzzy searches, and name variations. These matters also affect Ancestry’s shaky leaf, an algorithm that tries to find relevant hints to match your ancestor’s profile. It is much like a manual index search of a database; however, instead, the computer automatically searches for you. When it does work, the hint function can yield information you may not have considered in your search.
Although the shaky leaf hints are often false leads, you can try to make that leaf work for you if you aren’t receiving any relevant hints. Try inserting information in a particular ancestor’s profile, such as a birth date or parent that another Ancestry member claims to be true, although she may have no reference or proof of the fact. Or you may have a reason to believe your ancestor died in a particular year. Sometimes inserting unproven information does nothing, but other times I have had good luck with this method.
You can also try various spellings of your ancestor’s name. For example, I was searching for more information on a Jacob Kerwer in the early 1800s, a German name that was usually Anglicized as Karver or Carver. My family tree profile had him listed as Kerwer, which didn’t yield any shaky leaf hints, so I changed his name to Karver and found many hints I could sift through, a few of which were relevant. When I was finished, I set his name back to the German spelling of Kerwer. As I have pointed out in the past, many surnames are quite varied, such as Jaeger becoming translated as Hunter or O' Maolchaoin becoming Anglicized as Mulqueen. Ancestry’s algorithms can’t differentiate between such diversity of names.
A word of advice—If you are including unproven information in an ancestor profile, whether the tree is public or private, make a note of it for yourself and/or others to clearly see. An example of my three-times great grandfather William Corbit’s profile looks like the graphic below:
Always experiment with a database. Use many variations in index searching, and try to find clues by manipulating profile information. And always question its recall of information. After all, it is only a computer program that is bound to mislead you if you don’t take control of it.
Next week we will take a look at race, color, and ethnicity.
