Joan Silverman wrote:

"Here is my dilemma . . . One of my great uncles, Israel F. Goldberg was
killed in action in WWI body never recovered. I thought I had finally found
his birth certificate, however, it came back 4 years off. According to the
1892, 1900, 1910 federal census, and 1905 NY census, and his draft card he
was born in 1888. However his birth certificate states January 1892 and is
stamped February 1892. I do believe this is his certificate because of his
mother and father's name, father's occupation and the street they were
living on at the time are all correct. I realize that being one of 11,
although he was one of 7 at the time, and so long ago, people couldn't
remember when they were born but this has happened frequently with this
large family. Which birthdate should I use?"

The issue of different and even widely disparate birth dates is very common
in genealogical research. This is a good example of how careful evaluation
will help you resolve conflicting evidence. Not all sources of evidence are
equal, and in this case I would take the birth certificate to be most
accurate, given that his parents' names, father's occupation, and residence
are confirmed by other sources, and 1892 is not so far removed >from the
dates on the other sources to be unlikely. Important factors to consider
include the proximity to the event being recorded, in this case the birth,
the informant, and whether the informant was likely to be an eye witness
and, unlike the infant, cognizant of the date.

Elizabeth Shown Mills is the leading professional in the area of evidence
evaluation and has a wonderful website with lots of free information and a
link to purchasing her books and QuickSheets at www.evidenceexplained.com.
There is also a discussion forum that you can register for on the site to
learn more and ask questions.

Lesley K. Cafarelli
Minneapolis, MN, USA