AP Statistics / Mr. Hansen |
Name: _________________________ |
Answers
to “Stan Stanford” Questions
1. |
If X
= # who say yes, then E(X) » np = 30(0.045) = 1.35 females. |
|
|
2. |
Neither, since the condition
of independence is not satisfied.
(An SRS is sampling without
replacement, which does not give independent trials.) However, since the
population (300) is at least 10 times larger than the sample (30), we neglect
the lack of independence. The other conditions of the binomial setting are
satisfied, namely §
X counts
the number of successes §
n is fixed
(n = 30) §
2 possible
outcomes (acceptance or rejection) on each trial Therefore, we will use B(30, 0.045) as a reasonable
approximation of the true distribution. |
|
|
3. |
sX = Ö(npq) = Ö(30 · 0.045 · 0.955) = 1.135 females |
|
|
4. |
If Y
= # asked in order to obtain first acceptance, then E(Y) » 1/p =
1/0.045 = 22.222 females. |
|
|
5. |
Neither, since as before, independence is violated.
However, the trials are approximately independent since the population is
large. (Note that we cannot compare the population to n this time, since n is
undefined for the geometric setting.) The other conditions of the geometric
setting are satisfied, namely §
Y counts
the number of trials needed in order to obtain the first success §
n is
unspecified §
2 possible
outcomes (acceptance or rejection) on each trial Therefore, we will use the geometric
distribution with p = 0.045 (which
we might call G(0.045),
though that notation is not standardized) as a reasonable approximation of
the true distribution. |
|
|
6. |
Select pseudorandom integers between 000
and 999 inclusive, where 000 through 044 denote success and everything else denotes
failure. Repeated values are acceptable. Since these are independent trials,
the simulation does not faithfully capture the conditions of the problem.
However, as noted above, we will neglect the lack of independence in the
original problem, since the population is large. For each run of random
integers, record how many had to be selected in order to obtain one that was
044 or less. (For example, if success occurred on the 12th integer, record
“12” on the tally sheet.) Compute statistics after 20 values have been
recorded in a data set. |
|
|
7. |
Sample s.d. = 14.657,
an estimate for sY |
|
|
8. |
The simulation produced an estimate for E(Y)
that was low, more than 4 females below the theoretical value of 22.222. |
|
|
9. |
By formula on p. 475 of main text, sY = Ö(q/p2) = Ö(0.955/0.0452) = 21.716. The simulation produced an
estimate for sY that was more than 7 females too low. |