Running Head: WTVX CASE STUDY 1

WTVX Case Study

American Public University System

September 19, 2010

1. What are the chances of getting 15 days of rain during the next 30 days?

In this case study, the question calls for the probability of a specific number of successes, 15

days of rain, out of a set number of trials, the next 30 days. This question calls for the Bernoulli process.

The Bernoulli process is described by the binomial probability distribution which has four requirements

(Render, B., Stair, R.M., & Hanna, M.E., 2009, p. 41).

a. Each trial must have only two possible outcomes. The two possible outcomes are rain or
shine.
b. The probability stays the same from one trial to the next. Joe’s forecast “was that there
would be a 70% chance of rain every day” (Render, B., Stair, R.M., & Hanna, M.E., 2009, p.
65).
c. The trials are statistically independent. Joe points out “what happens on one day (rain or
shine) was not in any way dependent on what happened the day before” (Render, B., Stair,
R.M., & Hanna, M.E., 2009, p. 65).
d. The number of trials is a positive integer. In this case, there are 30 trials.

If the problem does not meet the four requirements above, the binomial probability distribution cannot

be used. Since WTVX meets all of the requirements, as indicated above, the binomial probability

distribution can be used. The formula for the probability of r successes in n trials is

n!
p r qn−r
r ! ( n−r ) !

Constant Definition This case study

Plugging in the numbers from the chart above, the probability of 15 days of rain in 30 days is

30 !
.7 15 .330−15 = .01057 or 1.057%
15! ( 30−15 ) !
WTVX CASE STUDY 3

This is a long equation because the symbol ! means factorial which equals n(n-1)(n-2)….(1). However, it

can be done in excel by using the BINOMDIST formula. When using this formula you merely need to

input the numbers from the chart above, and excel does the work for you.

In some cases, you can also use a binomial distribution chart. Quantitative Analysis for

Management has a chart in Appendix B (Render, B., Stair, R.M., & Hanna, M.E., 2009, p. 722-26). This

chart is not sufficient here, because it only provides results for up to 20 trials and this case study asks for

the probability in 30 days or trials. However, assuming we wanted to know the probability of 15 days of

rain in 20 days with all other factors the same, the chart can be used. You merely locate n=20 on the left

side of the chart, then move down to 15 as the number of successes under 20. Finally, move across to

the column that indicates .70 as the probability of each event. The solution would be .1789.

2. What do you think about Joe’s assumptions concerning the weather for the next 30 days?

The solution above and Joe makes two major assumptions: (a) that the probability stays the

same every day; and (b) that the trials are independent. I am not a weather expert, but if it rains today,

isn’t it less likely that it will rain tomorrow? And if it doesn’t rain for a whole week, doesn’t that increase

the probability that it will rain the next day? I believe that the probability does not stay the same

everyday because the trials are not independent. The probability of rain depends on the amount of

moisture in the air which is affected by whether it did or did not rain the day before. If Joe did not make

these assumptions, then to determine the probability of rain on one day would be dependent on the

previous day(s) events. If you do not accept these two assumptions, then you cannot use binomial

probability distribution to determine the probability.

The case study also makes the assumption that rain and shine are the only two options; that rain

and shine are collectively exhaustive. One could argue that this is untrue because a day can start rainy

and end sunny or vice versa. In that case, does the day count as rainy or sunny? Moreover, how much
WTVX CASE STUDY 4

rain is necessary for it to count as a rainy day, one drop in one spot? Further, couldn’t it snow?

Assuming the day is considered rainy if it rains for any part of the day and that any type of precipitation,

i.e. snow, is considered rain, you could still calculate the probability of rain on 15 of 30 days, but it

becomes a series of equations.

First, it would be necessary to establish the different combinations in which you would have rain

exactly 15 of 30 days. For instance, you could have rain the first 15 days straight and then have shine

the last 15 days, you could have the opposite of that, or you could have any other combination.

(Render, B., Stair, R.M., & Hanna, M.E., 2009, p. 41)

Once those possibilities are established, you have to find the conditional probability of rain on

each day taking those combinations into account. For instance, if it rains for 14 days straight the

probability that it will rain on the 15 th day will be different than if it was sunny for 14 days straight. This

P ( AB)
is the difficult part of the problem. Conditional probability is found using the equation
P(B)

(Render, B., Stair, R.M., & Hanna, M.E., 2009, p. 31). Finally, you would take each probability and

multiply them in order to find the overall probability of exactly 15 days of rain. This is joint probability of

dependent events.
WTVX CASE STUDY 5

References

Render, B., Stair Jr., R.M., & Hanna, M.E. (2009). Quantitative analysis for management. Upper Saddle
River, NJ: Pearson Prentice Hall.