Assignment for Monday, August 23, 1999
Complete the in-class activity we did on
Friday, as far as you can go without using ProbSim. That is,
State what assumptions you need to make about the situation.
For example: For #1 we assumed that birthdays are spread evenly
throughout the year and that people in rooms are there by a random
process that does not lead to groups whose birthdays are biased
toward a particular set of dates.
Describe your method of investigation.
In #1 we set up ProbSim to simulate selecting birthdates at
random for 25 people:
- We used the numbers 1 through 365 to represent possible birthdates
. We set up ProbSim so that it had 365 labels ("1"
through "365") and so that the mixer contained one
of each label.
- We set ProbSim to select samples WITH REPLACEMENT. This is
because when we select a birthday at random, that birthdate is
still available the next time we select another birthdate for
the same sample.
- We set ProbSim to select samples of size 25
- We set ProbSim to select 40 samples
- We analyzed the results by directing ProbSim to count all
those samples that had the UNORDERED pattern of
V1 V1 * * * * * * * * * * * * * * * * * * * * * * *
This means that we directed ProbSim to count the number of samples
that had at least two birthdates the same. (So, it counted 78
118 32 184 118 92 32 ... as one sample that had at least two
birthdates the same.)
You really can't do anything for this until you run the simulation.
But, it might be useful if you try to ANTICIPATE the results you'll