You are on page 1of 5

BANA 6610: Homework 2

1. The amount of time that a union stays on strike is judged to follow an


exponential distribution with a mean of 10 days.
(a) Find the probability that a strike lasts less than one day.
For all of these problem note that the rate is actually the inverse of
the mean. Also since we are dealing with a continuous districution,
we can find the probability of being below the number of days exactly.
If the distribution was discrete, we would actually use a number of
days of one day lower for the below cases.
> pexp(1,1/10)
[1] 0.09516258
(b) Find the probability that a strike lasts less than six days.

> pexp(6,1/10)
[1] 0.4511884
(c) Find the probability that a strike lasts between six and seven days.

> pexp(7,1/10) - pexp(6,1/10)


[1] 0.05222633
(d) Find the conditional probability that a strike lasts less than seven
days, given that it already lasted six days. Compare your answer to
part (a).

P (X<7X>6)
P (X < 7|X > 6) = P (X>6) =
> (pexp(7,1/10) - pexp(6,1/10))/(1 - pexp(6,1/10))
[1] 0.09516258
(e) Find the probability that a strike lasts between six and seven days.

> pexp(7,1/10) - pexp(6,1/10)


[1] 0.05222633
The answer to (a) is the same are the same.
(f) Plot the density to help someone understand the distribution of
strikes.

1
> x <- seq(35, 0, len = 100)
> curve(dexp(x, 1/10), from = 0, to = 35, xlab="Days",
+ main="Exponential Distribution", ylab="Density")

Exponential Distribution
0.00 0.04 0.08
Density

0 5 10 15 20 25 30 35
Days

2. In a city, 60% of the voters are in favor of building a new park. An


interviewer intends to conduct a survey.
(a) If the interviewer selects 20 people randomly, what is the probability
that more than 15 of them will favor building the park?

> 1- pbinom(15, 20, .6)


[1] 0.05095195
(b) Instead of choosing 20 people as in part (a), suppose that the inter-
viewer wants to conduct the survey until she has found exactly 12
who are in favor of the park. What is the probability that the first
12 people surveyed all favor the park (in which case the interviewer
can stop)? What is the probability that the interviewer can stop af-
ter viewing the thirteenth subject? What is the probability that the
interviewer can stop after interviewing the eighteenth subject?

Using the negative binomial distribution we have:


P (X = 13) = 12
 1 12
1 .4 .6 = 0.01
P (X = 18) = 176 .46 12
.6 = 0.11

2
3. The length of time until a strike is settled is distributed uniformly from 0
to 10.5 days.
(a) Find the probability that a strike lasts less than one day.

> punif(1,0, 10.5)


[1] 0.0952381
(b) Find the probability that a strike lasts less than six days.

> punif(6,0, 10.5)


[1] 0.5714286
(c) Find the probability that a strike lasts between six and seven days.

> punif(7,0, 10.5) - punif(6,0, 10.5)


[1] 0.0952381
(d) Find the conditional probability that a strike lasts less than seven
days, given that it already lasted six days. Compare your answer to
part (a).

P (X<7]X>6)
P (X < 7|X > 6) = P (X>6)

> (punif(7,0,10.5) - punif(6,0,10.5))/(1 - punif(6,0,10.5))


[1] 0.2222222
The answer is different than the answer in part (a) for the uniform
distribution.
(e) Plot the density to help someone understand the distribution of
strikes.

> x <- seq(12, -1, len = 100)


> curve(dunif(x, 0,10.5), from = -1, to = 12, xlab="Days",
+ main="Uniform Distribution", ylab="Density")

3
Uniform Distribution

0.00 0.04 0.08


Density

0 2 4 6 8 10 12
Days

4
4. You are an executive at Proctor and Gamble and are about to introduce a
new product. Your boss has asked you to predict the market share (X, a
proportion between 0 and 1) that the new product will caputre. You are
unsure of X, but have found the best model to illustrate the distribution
of market share is a beta distribution with an = 2 and = 5.

(a) What is P (0.08 < X < 0.14)?

> pbeta(.14,2,5) - pbeta(.08, 2, 5)


[1] 0.1229882
(b) What is P (0.14 < X < 0.22)?

> pbeta(.22,2,5) - pbeta(.14, 2, 5)


[1] 0.1934193
(c) What is the amount of market share you can expect to own?
2
+ = 7
(d) The boss suggests that if the market share is less than 15%, then the product
should not be launched. What is the probability that the product earns less
than 15%?

> pbeta(.15,2,5)
[1] 0.2235157
(e) What two values of market share can you expect to be between 95% of the time?

> qbeta(.025,2,5)
[1] 0.04327187
> qbeta(.975,2,5)
[1] 0.6412346
(f) Based on your answer to the previous questions, would you suggest launching
the product? Explain.
There is a 22.35% chance that the market share will be less than 15%.
There is a 95% chance the market share will be between 4.3% and 22.35%.
Given these statistics, I would not suggest launching.

You might also like