Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: Marilyn, again

On Mon, 4 Aug 1997, John Mallinckrodt wrote:

Marilyn: Take 2

As Marilyn and the rest of us all know, it takes only 23 randomly
chosen people to have a greater than 50% chance that at least two
will have the same birthday. Yesterday Marilyn published a letter
from a reader who claimed that, for a group of 50 people, the
result becomes "virtually certain." She took the reader to task
for falling victim to a "persistent, erroneous extrapolation" of
the preceding fact.

So, here we go again. You can calculate the probability yourself.
What do you think of Marilyn's answer?

Semantics mostly I'm sure. I have two first hand experiences with doing
the birthday test on a group of this approximate size. The first one
found a few double birthdays and at least one triple. The second found
none at all. A small sample size to be sure, but enough to make me wary
of the term "virtually certain".

Unless somebody has a table of birthday:date probabilities which is
appropriate for whatever population is being sampled, the answer to
this question is not the simple one we calculate anyways.

|++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++|
| Doug Craigen |
| |
| If you think Physics is no laughing matter, think again .... |
| http://cyberspc.mb.ca/~dcc/phys/humor.html |
|++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++|