4.2 The Rules of Probability
|
Chapter 4 1654-22. Several years ago, a large state university conducted a survey of undergraduate students regarding their use of computers. The results of the survey are contained in the data file ComputerUse.
a. Based on the data from the survey, what was the probability that undergraduate students at this university will have a major that requires them to use a computer on a daily basis?
b. Based on the data from this survey, if a student was a business major, what was the probability of the student believing that the computer lab facilities are very adequate?
c. Select a sample of 20 students at your university and ask them what their major is and whether the major requires them to access the Internet for computational purposes, with a personal computer, notebook, tablet, or other device. Use these data to calculate the probability of each major requiring a computing device, and then compare your results with those from the older survey.
4-23. A scooter manufacturing company is notified whenever a scooter breaks down, and the problem is classified as being either mechanical or electrical. The company then matches the scooter to the plant where it was assembled. The file Scooters contains a random sample of 200 breakdowns. Use the data in the file and the relative frequency assessment method to find the following probabilities:
a. What is the probability a scooter was assembled at the Tyler plant?
b. What is the probability that a scooter breakdown was due to a mechanical problem?
c. What is the probability that a scooter was assembled at the Lincoln plant and had an electrical problem?
4-24. A survey on cell phone use asked, in part, what was the most important reason people give for not using a wireless phone exclusively. The responses were:
(1) Like the safety of traditional phone, (2) Like having a land line for a backup phone, (3) Pricing not
attractive enough, (4) Weak or unreliable cell signal at home, (5) Need phone line for DSL Internet access, and (6) Other. The file titled Wireless contains the responses for the 1,088 respondents.
a. Calculate the probability that a randomly chosen respondent would not use a wireless phone exclusively because of some type of difficulty in placing and receiving calls with a wireless phone.
b. Calculate the probability that a randomly chosen person would not use a wireless phone exclusively because of some type of difficulty in placing and receiving calls with a wireless phone and is over the age of 55.
c. Determine the probability that a randomly chosen person would not use a wireless phone exclusively because of a perceived need for Internet access and the safety of a traditional phone.
d. Of those respondents under 36, determine the probability that an individual in this age group would not use a wireless phone exclusively because of some type of difficulty in placing and receiving calls with a wireless phone.
4-25. A selection of soft-drink users is asked to taste two disguised soft drinks and indicate which they prefer.
The file titled Challenge contains the results of a study that was conducted on a college campus.
a. Determine the probability that a randomly chosen student prefers Pepsi.
b. Determine the probability that one of the students prefers Pepsi and is less than 20 years old.
c. Of those students who are less than 20 years old, calculate the probability that a randomly chosen student prefers (1) Pepsi and (2) Coke.
d. Of those students who are at least 20 years old, calculate the probability that a randomly chosen student prefers (1) Pepsi and (2) Coke.
The Rules of Probability
166 Chapter 4
|
Introduction to ProbabilityAll possible outcomes associated with an experiment form the sample space. Therefore, the sum of the probabilities of all possible outcomes is 1, as shown by Probability Rule 2.
Addition Rule for Individual Outcomes If a single event is made up of two or more individual outcomes, then the probability of the event is found by summing the probabilities of the individual outcomes. This is illustrated by Probability Rule 3.
Probability Rule 2
a
k i=1
P1ei2 = 1 where:
k = Number of outcomes in the sample ei = ith outcome
(4.4)
Probability Rule 3: Addition Rule for Individual Outcomes
The probability of an event Ei is equal to the sum of the probabilities of the individual outcomes that form Ei. For example, if
Ei = 5e1, e2, e36 then
P1Ei2 = P1e12 + P1e22 + P1e32 (4.5)
TABLE 4.2 Google’s Survey Results
Searches per Day Frequency Relative Frequency
e1 =At least 10 400 0.08
e2 = 3 to 9 1,900 0.38
e3 = 1 to 2 1,500 0.30
e4 = 0 1,200 0.24
Total 5,000 1.00
BUSINESS APPLICATION
Addition Rule
Google Google has become synonymous with Web searches and is the leader in the search engine marketplace. Suppose officials at the northern California headquarters have recently performed a survey of computer users to determine how many Internet searches individuals do daily using Google. Table 4.2 shows the results of the survey of Internet users.
The sample space for the experiment for each respondent is SS = 5e1, e2, e3, e46 where the possible outcomes are
e1 = At least 10 searches e2 = 3 to 9 searches e3 = 1 to 2 searches e4 = 0 searches Probability Rule 1
For any event Ei,
0 … P1Ei2 … 1 for all i (4.3)
o u tc o m e 2
M04_GROE0383_10_GE_C04.indd 166 23/08/17 7:07 PM
4.2 The Rules of Probability
|
Chapter 4 167EXAMPLE 4-7
The Addition Rule for Individual Outcomes
KQRT 1340 Radio The KQRT 1340 radio station is a combination news/talk and “oldies” station. During a 24-hour day, a listener can tune in and hear any of the following four programs being broadcast:
“Oldies” music News stories Talk programming Commercials
Recently, the station has been having trouble with its transmitter. Each day, the sta- tion’s signal goes dead for a few seconds; it seems that these outages are equally likely to occur at any time during the 24-hour broadcast day. There seems to be no pattern regard- ing what is playing at the time the transmitter problem occurs. The station manager is concerned about the probability that these problems will occur during either a news story or a talk program.
s t e p 1 Define the experiment.
The station conducts its broadcast starting at 12:00 midnight, extending until a transmitter outage is observed.
s t e p 2 Define the possible outcomes.
The possible outcomes are each type of programming that is playing when the transmitter outage occurs. There are four possible outcomes:
e1 = Oldies e2 = News
e3 = Talk programs e4 = Commercials
s t e p 3 Determine the probability of each possible outcome.
The station manager has determined that out of the 1,440 minutes per day, 540 minutes are oldies, 240 minutes are news, 540 minutes are talk programs, and 120 minutes are commercials. Therefore, the probability of each type of programming being on at the moment the outage occurs is assessed as follows:
Using the relative frequency assessment approach, we assign the following probabilities:
P1e12 = 400>5,000 = 0.08 P1e22 = 1,900>5,000 = 0.38 P1e32 = 1,500>5,000 = 0.30 P1e42 = 1,200>5,000 = 0.24 g = 1.00
Assume we are interested in the event respondent performs 1 to 9 searches per day:
E = Internet user performs 1 to 9 searches per day The outcomes that make up E are
E = 5e2, e36
We can find the probability, P(E), by using Probability Rule 3 (Equation 4.5), as follows:
P1E 2 = P1e22 + P1e32
= 0.38 + 0.30
= 0.68
M04_GROE0383_10_GE_C04.indd 167 23/08/17 7:07 PM
168 Chapter 4
|
Introduction to ProbabilityOutcome P1ei2
e1 = Oldies
P(e1) = 540
1,440 = 0.375 e2 = News
P(e2) = 240
1,440 = 0.167 e3 = Talk programs
P(e3) = 540
1,440 = 0.375 e4 = Commercials
P(e4) = 120
1,440 = 0.083 a = 1.000
Note that based on Equation 4.4 (Probability Rule 2), the sum of the probabilities of the individual possible outcomes is 1.0.
s t e p 4 Define the event of interest.
The event of interest is a transmitter problem occurring during a news or talk program. This is
E = 5e2, e36
s t e p 5 Use Probability Rule 3 (Equation 4.5) to compute the desired probability.
P1E 2 = P1e22 + P1e32 P1E 2 = 0.167 + 0.375 P1E 2 = 0.542
Thus, the probability is slightly higher than 0.5 that when a transmitter problem occurs, it will happen during either a news or talk program.
TRY EXERCISE 4-26 (pg. 184)
Complement Rule Closely connected with Probability Rules 1 and 2 is the complement of an event. The complement of event E is represented by E. The Complement Rule is a corollary to Probability Rules 1 and 2.
Complement
The complement of an event E is the collection of all possible outcomes not contained in event E.
Complement Rule
P1E 2 = 1 - P1E 2
That is, the probability of the complement of event E is 1 minus the probability of event E.
(4.6)
EXAMPLE 4-8
The Complement Rule
Capital Consulting The managing partner for Capital Consulting is working on a proposal for a consulting project with a client in Sydney, Australia. The manager lists four possible net profits from the consulting engagement and his subjectively assessed proba- bilities related to each profit level.
Outcome P(Outcome)
$ 0 0.70
$ 2,000 0.20
$15,000 0.07
$50,000 0.03
g = 1.00
M04_GROE0383_10_GE_C04.indd 168 23/08/17 7:07 PM
4.2 The Rules of Probability
|
Chapter 4 169Note that each probability is between 0 and 1 and that the sum of the probabilities is 1, as required by Probability Rules 1 and 2.
The manager plans to submit the proposal if the consulting engagement will have a positive profit, so he is interested in knowing the probability of an outcome greater than
$0. This probability can be found using the Complement Rule with the following steps:
s t e p 1 Determine the probabilities for the outcomes.
P1+02 = 0.70 P1+2,0002 = 0.20 P1+15,0002 = 0.07 P1+50,0002 = 0.03
s t e p 2 Find the desired probability.
Let E be the consulting outcome event =+0. The probability of the $0 outcome is
P1E2 = 0.70
The complement, E, is all investment outcomes greater than $0. From the Complement Rule, the probability of profit greater than $0 is
P1Profit 7 +02 = 1 - P1Profit = +02 P1Profit 7 +02 = 1 - 0.70
P1Profit 7 +02 = 0.30
Based on the manager’s subjective probability assessment, there is a 30%
chance the consulting project will have a positive profit.
TRY EXERCISE 4-32 (pg. 184)
Addition Rule for Any Two Events
BUSINESS APPLICATION
Addition Rule
Google (continued ) Suppose the staff who conducted the survey for Google discussed ear- lier also asked questions about the computer users’ ages. The Google managers consider age important in designing their search engine methodologies. Table 4.3 shows the breakdown of the sample by age group and by the number of times a user performs a search each day.
Table 4.3 shows that seven events are defined. For instance, E1 is the event that a com- puter user performs ten or more searches per day. This event is composed of three individual outcomes associated with the three age categories. These are
E1 = 5e1, e2, e36
In another case, event E5 corresponds to a survey respondent being younger than 30 years of age. It is composed of four individual outcomes associated with the four levels of search activity. These are
E5 = 5e1, e4, e7, e106
o u tc o m e 2
TABLE 4.3 Google Search Study
Age Group
Searches per Day E5
Under 30 E6
30 to 50 E7
Over 50 Total E1 Ú 10 searches e1 200 e2 100 e3 100 400 E2 3 to 9 searches e4 600 e5 900 e6 400 1,900 E3 1 to 2 searches e7 400 e8 600 e9 500 1,500 E4 0 searches e10 700 e11 500 e12 0 1,200
Total 1,900 2,100 1,000 5,000
M04_GROE0383_10_GE_C04.indd 169 23/08/17 7:08 PM
170 Chapter 4
|
Introduction to ProbabilitySuppose we wish to find the probability of E4 (0 searches) or E6 (being in the 30-to-50 age group). That is,
P1E4 or E62 = ?
TABLE 4.4 Google—Joint Probability Table
Age Group
Searches per Day E 5 Under 30 E 6 30 to 50 E 7 Over 50 Total
E1 Ú 10 searches e1 200>5,000 = 0.04 e2 100>5,000 = 0.02 e3 100>5,000 = 0.02 400>5,000 = 0.08 E2 3 to 9 searches e4 600>5,000 = 0.12 e5 900>5,000 = 0.18 e6 400>5,000 = 0.08 1,900>5,000 = 0.38 E3 1 to 2 searches e7 400>5,000 = 0.08 e8 600>5,000 = 0.12 e9 500>5,000 = 0.10 1,500>5,000 = 0.30 E4 0 searches e10 700>5,000 = 0.14 e11 500>5,000 = 0.10 e12 0>5,000 = 0.00 1,200>5,000 = 0.24 Total 1,900>5,000 = 0.38 2,100>5,000 = 0.42 1,000>5,000 = 0.20 5,000>5,000 =1.00
To find this probability, we must use Probability Rule 4.
The key word in knowing when to use Rule 4 is or. The word or indicates addition. Figure 4.1 is a Venn diagram that illustrates the application of the Addition Rule for Any Two Events.
Notice that the probabilities of the outcomes in the overlap between the two events, E1 and E2, are double-counted when the probabilities of the outcomes in E1 are added to those in E2. Thus, the probabilities of the outcomes in the overlap, which is E1 and E2, need to be sub- tracted to avoid the double-counting.
Probability Rule 4: Addition Rule for Any Two Events E1 and E2
P1E1 or E22 = P1E12 + P1E22 - P1E1 and E22 (4.7)
E1 E2
E1and E2
P(E1or E2) = P(E1) + P(E2) – P(E1and E2) FIGURE 4.1 Venn Diagram—
Addition Rule for Any Two Events
Table 4.3 illustrates two important concepts in data analysis: joint frequencies and mar- ginal frequencies. Joint frequencies, which were discussed in Chapter 2, are the values inside the table. They provide information on age group and search activity jointly. Marginal fre- quencies are the row and column totals. These values give information on only the age group or only Google search activity.
For example, 2,100 people in the survey are in the 30- to 50-year age group. This column total is a marginal frequency for the age group 30 to 50 years, which is represented by E6. Now notice that 600 respondents are younger than 30 years old and perform three to nine searches a day. The 600 is a joint frequency whose outcome is represented by e4.
Table 4.4 shows the relative frequencies for the data in Table 4.3. These values are the probability assessments for the events and outcomes.
M04_GROE0383_10_GE_C04.indd 170 23/08/17 7:08 PM
4.2 The Rules of Probability
|
Chapter 4 171Referring to the Google situation, the probability of E4 (0 searches) or E6 (being in the 30-to-50 age group) is
P1E4 or E62 = ?
Table 4.5 shows the relative frequencies with the events of interest shaded. The overlap cor- responds to the joint occurrence (intersection) of conducting 0 searches and being in the 30-to-50 age group. The probability of the outcomes in the overlap is represented by P1E4 and E62 and must be subtracted. This is done to avoid double-counting the probabilities of the outcomes that are in both E4 and E6 when calculating the P1E4 or E62. Thus,
P1E4 or E62 = P1E42 + P1E62 - P1E4 and E62
= 0.24 + 0.42 - 0.10
= 0.56
Therefore, the probability that a respondent will either be in the 30-to-50 age group or per- form 0 searches on a given day is 0.56.
What is the probability a respondent will perform 1 to 2 searches or be in the over-50 age group? Again, we can use Probability Rule 4:
P1E3 or E72 = P1E32 + P1E72 - P1E3 and E72 Table 4.6 shows the relative frequencies for these events. We have
P1E3 or E72 = 0.30 + 0.20 - 0.10 = 0.40
Thus, there is a 0.40 chance that a respondent will perform 1 to 2 searches or be in the over- 50 age group.
TABLE 4.5 Google Searches—Addition Rule Example
Age Group
Searches per Day E5 Under 30 E6 30 to 50 E7 Over 50 Total
E1 Ú 10 searches e1 200>5,000 = 0.04 e2 100>5,000 = 0.02 e3 100>5,000 = 0.02 400>5,000 = 0.08 E2 3 to 9 searches e4 600>5,000 = 0.12 e5 900>5,000 = 0.18 e6 400>5,000 = 0.08 1,900>5,000 = 0.38 E3 1 to 2 searches e7 400>5,000 = 0.08 e8 600>5,000 = 0.12 e9 500>5,000 = 0.10 1,500>5,000 = 0.30 E4 0 searches e10 700>5,000 = 0.14 e11 500>5,000 = 0.10 e12 0>5,000 = 0.00 1,200>5,000 = 0.24 Total 1,900>5,000 = 0.38 2,100>5,000 = 0.42 1,000>5,000 = 0.20 5,000>5,000 = 1.00
EXAMPLE 4-9
Addition Rule for Any Two Events
Greenfield Forest Products Greenfield Forest Products manufactures lumber for large material supply centers like Home Depot and Lowe’s in the United States and Canada. A representative from Home Depot is due to arrive at the Greenfield plant for a meeting to dis- cuss lumber quality. When the Home Depot representative arrives, he will ask Greenfield’s
TABLE 4.6 Google—Addition Rule Example
Age Group
Searches per Day E5 Under 30 E6 30 to 50 E7 Over 50 Total
E1 Ú 10 searches e1 200>5,000 = 0.04 e2 100>5,000 = 0.02 e3 100>5,000 = 0.02 400>5,000 = 0.08 E2 3 to 9 searches e4 600>5,000 = 0.12 e5 900>5,000 = 0.18 e6 400>5,000 = 0.08 1,900>5,000 = 0.38 E3 1 to 2 searches e7 400>5,000 = 0.08 e8 600>5,000 = 0.12 e9 500>5,000 = 0.10 1,500>5,000 = 0.30 E4 0 searches e10 700>5,000 = 0.14 e11 500>5,000 = 0.10 e12 0>5,000 = 0.00 1,200>5,000 = 0.24 Total 1,900>5,000 = 0.38 2,100>5,000 = 0.42 1,000>5,000 = 0.20 5,000>5,000 = 1.00
M04_GROE0383_10_GE_C04.indd 171 23/08/17 7:08 PM
172 Chapter 4
|
Introduction to Probabilitymanagers to randomly select one board from the finished goods inventory for a quality check. Boards of three dimensions and three lengths are in the inventory. The following table shows the number of boards of each size and length:
Dimension
Length E2 = 2″ * 4″ E5 = 2″ * 6″ E6 = 2″ * 8″ Total
E1 = 8 feet 1,400 1,500 1,100 4,000
E2 = 10 feet 2,000 3,500 2,500 8,000
E3 = 12 feet 1,600 2,000 2,400 6,000
Total 5,000 7,000 6,000 18,000
The Greenfield manager will be selecting one board at random from the inventory to show the Home Depot representative. Suppose he is interested in the probability that the board selected will be 8 feet long or a 2″ * 6″. To find this probability, he can use the following steps:
s t e p 1 Define the experiment.
One board is selected from the inventory and its dimensions are obtained.
s t e p 2 Define the events of interest.
The manager is interested in boards that are 8 feet long.
E1 = 8 feet
He is also interested in the 2″ * 6″ dimension, so E5 = 2″ * 6″ boards
s t e p 3 Determine the probability for each event.
There are 18,000 boards in inventory and 4,000 of these are 8 feet long, so P1E12 = 4,000
18,000 = 0.2222 Of the 18,000 boards, 7,000 are 2″ * 6″, so the probability is
P1E52 = 7,000
18,000 = 0.3889
s t e p 4 Determine whether the two events overlap, and if so, compute the joint probability.
Of the 18,000 total boards, 1,500 are 8 feet long and 2″ * 6″. Thus the joint probability is
P1E1 and E52 = 1,500
18,000 = 0.0833
s t e p 5 Compute the desired probability using Probability Rule 4.
P1E1 or E52 = P1E12 + P1E52 - P1E1 and E52 P1E1 or E52 = 0.2222 + 0.3889 - 0.0833
= 0.5278
The chance of selecting an 8-foot board or a 2″ * 6″ board is just under 0.53.
TRY EXERCISE 4-28 (pg. 184)
Addition Rule for Mutually Exclusive Events We indicated previously that when two events are mutually exclusive, both events cannot occur at the same time. Thus, for mutu- ally exclusive events,
P1E1 and E22 = 0
Therefore, when you are dealing with mutually exclusive events, the Addition Rule assumes a different form, shown as Probability Rule 5.
M04_GROE0383_10_GE_C04.indd 172 23/08/17 7:08 PM
4.2 The Rules of Probability
|
Chapter 4 173Figure 4.2 is a Venn diagram illustrating the application of the Addition Rule for Mutually Exclusive Events.