Chapter 4
• Discrete Probability Distributions
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS I. INTRODUCTION TO RANDOM VARIABLES AND PROBABILITY DISTRIBTIONSA. Random Variable x.
1. x represents a numerical outcome which could be a count or a
measurement resulting from a statistical experiment.a) For example:
Take a multiple choice test with 20 questions. Your raw score will be the number of
correct responses.1)
x = number of correct responses.2)
We can list the values for x: 0, 1, 2, 3,….20.3)
A possible value for the random variable is x = 18. b) For example:
Measure the mileage you drive between your home and your job.
1)x = number of miles you drive
2)We cannot list ALL of the possible responses here, since all
decimal numbers are possible.
3)A possible value for the random variable is x = 26.835 miles.
2. The term random is used because the particular value it takes in one
trial of an experiment occurs by chance.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS
B. Types of random variables.1. Discrete Random Variables.
a) Values can be listed
b) Usually determined by a counting process.
c) A discrete random variable takes on only certain values.
d) The only values are usually integers, however, an exception would
be a shoe size, which could include half sizes. However, a shoe size
would not be 7.23.2. Continuous Random Variables
a) Possible values cannot be put in a list since there are infinitely
many.b) Usually
determined by a measuring process.c) All measures in an
interval (including decimals and fractions) are candidates for replacement.
d) No values are skipped.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS
C. Probability distribution for a discrete random variable1. List each value x that may occur in the
experiment and assign the probability of its occurrence P(x).
a) P(x) = number of times x has occurred divided by total number
of times the experiment is performedb) The sum of all the
simple probabilities is 1. ∑ P(x) = 1c) The expected
value of a discrete probability distribution (also called the mean) is denoted μ.
1)μ = ∑xP(x)
a. To calculate μ, make a table.
1. List the values of x in a vertical column I.
2. List the corresponding probabilities in vertical
column II.
3. In vertical column III, calculate the product of
each value of x times its probability.
4. The expected value of x is the sum of the
products listed in column III.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS
d) The standard deviation of an probability distribution is denoted σ.
1)Since we use the same table in which we have calculated the
expected value, columns I and II already are formed and have x and P(x)
entries.2)
Subtract the mean μ from each value of x, thus computing the
deviations (x – μ). Form a vertical column IV for these deviations.
3)Square each deviation and form vertical column V for
(x – μ)2, the deviations squared.4)
Multiply each deviation squared by the corresponding
probability from column II.
Put these products into vertical column VI.5)
Add the entries in column VI.6)
Find the square root of the sum of column VI.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS
D. Probability distribution of a continuous random variable.1. At this time we will not be required to
evaluate these probabilities, but will deal with them in later chapters.
Section 4-1 Example: Page 203 #30Cats 0 1 2 3 4 5
Households 1941 349 203 78 57 40
a) Construct a frequency distribution
Find the total number of experimentsDivide the frequency of each x by the total.
x Freq/Total P(x)
0
1
2
3
4
5
.728
.029
.076
.131
.021
.015
1941/2668
40/2668
203/2668
78/2668
57/2668
349/2668
CHECK!!!Are all the P(x) values between 0 and 1?Do all the P(x) values add up to 1? (Or within a hundredth or so of 1?)
1.000
2668
Section 4-1 Example: Page 203 #30Cats 0 1 2 3 4 5
Households 1941 349 203 78 57 40
b) Find the mean.Multiply each x value by its probability.
x P(x) x P(x)
0
1
2
3
4
5
.728
.029
.076
.131
.021
.015
Add these up to get the mean (µ)
.000
.087
.152
.131
.084
.075
The Round Off Rule (Used for the Mean, Variance, & Standard Deviation)Round to one decimal place more than is used for the x-values.
.529
Since these x values were all whole numbers, we round to the nearest tenth.The mean is .5
x =x =
x =
x =
x =x =
Section 4-1 Example: Page 203 #30Cats 0 1 2 3 4 5
Households 1941 349 203 78 57 40
c) Find the variance.
x P(x) x-µ (x-µ)2 P(x)(x-µ)2
0 .728
1 .131
2 .076
3 .029
4 .021
5 .015
Subtract the mean from each x value (remember that the mean is .5)
0 – 0.5
1 - 0.5
-0.5
0.5
2 - 0.5
3 - 0.5
4 - 0.5
5 - 0.5
1.5
2.5
3.5
4.5
Square the answers of (x - µ) to make all values positive.
(-.5)2
(.5)2
(1.5)2
(2.5)2
(3.5)2
(4.5)2
.25
.25
20.25
2.25
6.25
12.25
(.728)(.25)
(.131)(.25)
(.076)(2.25)
(.029)(6.25)
(.021)(12.25)
(.015)(20.25)
.182
.181
.304
.171
.257
.033
Multiply the probability of each x by the square of x - µ -- (P(x)(x-µ)2)The sum of this column (rounded to the nearest 10th) is the variance.
1.128
The variance is 1.1
Section 4-1 Example: Page 203 #30d) Find the standard deviation.
The standard deviation is the square root of the variance.
The square root of 1.1 is 1.05, which rounds to 1.1 (remember the round off rule).So, in summary, the mean was .5, the variance was 1.1, and the standard deviation was 1.1What this means is that in this one small town, the average household has .5 cats, with a standard deviation of 1.1.
𝜎=√𝜎2
Identify each of the following as either a discrete or continuous random variable.A. The number of people who are in a car B. The number of miles you drive in one week. C. Weight of a box of cereal. D. The number of boxes of cereal you buy in a year. E. Length of time you spend for lunch. F. The number of patients on a psychiatric ward in one day. G. The volume of blood which is transfused during an operation.
Discrete -- countable
Continuous -- measured
Continuous -- measured
Discrete -- countable
Continuous -- measured
Discrete -- countable
Continuous -- measured
In a personality inventory test for passive-aggressive traits, the possible scores are 1= extremely passive; 2 = moderately passive; 3= neither; 4 = moderately aggressive; and 5 = extremely aggressive. The test was administered to a group of 110 people and the results were as follows:
Construct a probability distribution table and compute the expected value (the mean) and the standard deviation.
X (score) 1 2 3 4 5F (Freq.) 19 23 32 26 10
x f P(x) x * P(x) x - µ (x - µ)2 (x - µ)2 * P(x)
1 19
2 23
3 32
4 26
5 10
Σ 110
X (score) 1 2 3 4 5F (Freq.) 19 23 32 26 10
.173
.209
.291
.236
.091
1.000
.173
.418
.873
.944
.455
2.863 = 2.9
-1.9 3.61 .625
-.9 .81 .169
.1 .01 .003
1.1 1.21 .286
2.1 4.41 .401
1.484 =1.5
The Mean (Expected Value) is 2.9, and the Variance is 1.5Remember that the standard deviation the square root of the
variance.The standard deviation is 1.2
Your assignment is pages 201 – 203, # 9 - 33 Odd
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2 I. BINOMIAL PROBABILITIES
In the previous chapter, we looked at statistical experiments and computed the probabilities of specified events. We will now examine a particular type of statistical experiment called a binomial experiment.
A. Characteristics of a binomial experiment.1. The same action is
repeateda)
Conditions for repetition must be identical.b)
One trial must be independent of all others. (the results of one trial
cannot affect another.)c)
The number of trials is n.2. A binomial
experiment must have exactly two outcomes.a)
Success (defined in the problem)b)
Failure (all outcomes that do not qualify as successes.)3. On an individual
triala)
P (success) = pb)
P (failure) = qc)
q = 1 – p (They are complementary events)
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2
4. The number of trials that are successful is denoted r where r ≤ n.
B. The basic problem is to find the probability of getting exactly r successes out
of n trials1. In a binomial
experiment we use r as the random variable.a)
r counts the number of successful outcomes.b)
r is a discrete random variable.c)
For each experiment, the possible values of r can be listed, r = 0, 1,
2, 3, … n. The values range from r = 0 (no successful outcomes) to r
= n (all outcomes are successful).2. The probability of
getting exactly 0 successes (all failures) is denoted P(0), the probability of getting exactly 1 success is
denoted P(1), and so on.a)
Since the categories of no successes, exactly one success, exactly
two successes, etc. are mutually exclusive events, we can add their
probabilities to answer a question using the OR combination.b)
The sum of all probabilities = 1. P(0) + P(1) + P(2) + P(3) + .. P(n) = 1.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2
C. Methods of determining the probability of r successes out of n trials.
1. Formula
a) P(r) = Cn,r pr qn-r
2. Table of Binomial Probabilities found in Appendix II of your text.
3. The TI-84 can solve a binomial distribution problem for us, so that’s the
way we are going to go.
a) Press 2nd VARS (DISTR).
b) This takes you to a screen that lists 16 different types of statistical
distributions. We will be returning to this screen A LOT!!
1) You are looking for the 11th and/or 12th type of distribution
listed.
a). Pressing the Alpha and A keys takes you to the
binompdf function.
1. binompdf gives you the probability that you will
receive PRECISELY (exactly) r successes.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2
b) Pressing the Alpha and B keys takes you to the binomcdf
function.
1.binomcdf gives you the CUMULATIVE probability for up
to r successes.
c) Once you have selected the type of distribution you want,
either binompdf or binomcdf, you enter the parameters in
identical fashion.
1. The calculator screen will prompt you for trials, p, and
x.
a. enter the values asked for, highlight Paste, and
press Enter.
b. Press the Enter key again to receive your answer.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2
D. THE MEAN AND STANDARD DEVIATION OF THE BINOMIAL DISTRIBUTION
1. The binomial distribution’s special properties allow for an easier formula
to compute the expected value (mean), the variance, and the standard
deviation.a.
To compute the mean of a binomial distribution, simply multiply
the number of trials by the probability of success on each trial.
Use b.
The variance of a binomial distribution is found by .
The standard deviation of a binomial distribution is 2. Application of the
binomial distributiona.
Useful in determining the probability of meeting certain quotas or
specifications under success/failure conditions.b.
For these problems, you’ll be determining the value of n (the
number of trials) necessary to meet the given quota.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2 EXAMPLE As an example, let’s go through Question Number 2 on Page 235.
A. n = 10 (number of questions)p = 1/5, or .20 (probability of guessing the
correct answer on one question)r = 10 for part a, 0 for part b, is greater than or
equal to 1 for part c, and greater than or equal to 5 for part d.
B. Let’s do part a.1. This is asking us
for the probability of getting PRECISELY 10 correct, so we use the binompdf.
2. 2nd VARS Alpha Math (A) gives us binompdf.
You can also just scroll until you see binompdf in the list.
3. Fill in the values 10,.2,10, highlight Paste and press Enter.
a. Your screen should have binompdf(10,.2,10) showing.4. Press Enter. The
screen will tell you that the probability of getting all ten questions correct is 1.024 x 10-7.
That means that you have a 0.0000001024 chance of getting all 10
correct.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2 EXAMPLE
C. Part b asks for the probability of getting none of them correct.
1. This is also asking for a precise probability, so we again go to binompdf.
2. This time, we enter 10,.2,0 into the calculator and get the answer of
.1073741824. That means that
you have a 10.74% chance of getting a zero on that quiz.D. Part c asks for the probability of getting at least one right.
This is a CUMULATIVE probability, as we need to add the probability of 1 to the probability of 2, all the way up to the probability of 10.
1. The smart way to do this is to realize that the complement of “at least
1” is “0”. Since we have already found P(0) to be .107, we can subtract
that from 1 and get the answer of .893
E. Part d asks for the cumulative probabilities for r = 5,6,7,8,9, and 10. The
calculator can’t do that, but it can find the probability of r = 0,1,2,3, and 4. This is the complement of 5 and up, so we can subtract this from 1 to get our
answer.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESDISCRETE PROBABILITY DISTRIBUTIONS SECTION 4-2 EXAMPLE
1.binomcdf(10,.2,4) gives the cumulative probabilities for r = 0,1,2,3, and
4. This
number is .9672. When
this is subtracted from 1, we get .0328. So, you
have about a 3% chance of getting at least a 50 on that quiz.Maybe
we should try studying!!F. The calculator will ALWAYS give you the
probability of less than or equal to the number you enter for x.
If you want greater than x, you need to use the complement rule.
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-2 EXAMPLES For each of the following:
State whether or not it is a binomial experiment.If it is a binomial experiment, describe “success” and “failure”.Identify values for n, q, and the range of values for r.
U.S.A. Today reported on July 27, 1990, that 70% of the people questioned said that they watched less T.V. than they did a year ago, 22% said they watched the same amount, and 8% said they watched more. Find the probability that exactly 3 out of a randomly selected group of 5 will say they watch less T.V. this year than last.
This a binomial experiment, with success being “watch less T.V. this year than last” and failure being “watch the same or more T.V. than last year”.
n = 5p = .70 (probability that one person will say they watch
less T.V. this year)q = .30 (1-.7)r = 3 (number of successes)
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-2 EXAMPLES For each of the following:
State whether or not it is a binomial experiment.If it is a binomial experiment, describe “success” and “failure”.Identify values for n, q, and the range of values for r.
There are 20 M&M candies in a dish. 8 are brown, 3 are red, 5 are green and 4 are yellow. Two candies are picked from the dish at random. What is the probability that both are red?
This is not a binomial experiment, since the trials are not independent.The probability that the first one is red is 3/20.The probability of the second being red, given that the first one was, is 2/19.
(3/20)*(2/19) = 6/380 = 3/190 = .0158
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-2 EXAMPLES For each of the following:
State whether or not it is a binomial experiment.If it is a binomial experiment, describe “success” and “failure”.Identify values for n, q, and the range of values for r.
A ten question multiple choice test is given. Each question has four choices. You did not study and have no clue as to any of the answers, so you have to randomly guess each answer. What is the probability you guess exactly 6 correctly (and pass)?
This is a binomial experiment.Success is “guessing the answer correctly” and failure is “guessing incorrectly”n = 10 (the number of questions)p = .25 (since there is one chance in four of guessing each answer correctly)q = .75 (1 - .25)r = 6
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-2 EXAMPLES For each of the following:
State whether or not it is a binomial experiment.If it is a binomial experiment, describe “success” and “failure”.Identify values for n, q, and the range of values for r.
U.S.A. Today reported on July 27, 1990, that 70% of the people questioned said that they watched less T.V. than they did a year ago, 22% said they watch the same amount, and 8% said they watch more. Find the probability that exactly 3 out of a randomly selected group of 5 will say they watch less T.V. this year than last.
On the calculator, use binompdf, since we want exactly 3 successes.2nd VARS binompdf, then enter 5, .7, and 3.Paste and Enter will give you .3087, which rounds to .309.)
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-2 EXAMPLES For each of the following:
State whether or not it is a binomial experiment.If it is a binomial experiment, describe “success” and “failure”.Identify values for n, q, and the range of values for r.
U.S.A. Today reported on July 27, 1990, that 70% of the people questioned said that they watched less T.V. than they did a year ago, 22% said they watch the same amount, and 8% said they watch more. Find the probability that between 2 and 4 (inclusive) people out of 5 will say they watched less T.V. this year than last year.
Use the binompdf function to find P(2), P(3), and P(4).2nd VARS binompdf (5, .7, 2), (5, .7, 3) and (5, .7, 4)
P(2) = .132P(3) = .309P(4) = .360
P(between 2 and 4 inclusive) = P(2) + P(3) + P(4) = .132 + .309 + .360 = .801
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-2 EXAMPLES For each of the following:
State whether or not it is a binomial experiment.If it is a binomial experiment, describe “success” and “failure”.Identify values for n, q, and the range of values for r.
U.S.A. Today reported on July 27, 1990, that 70% of the people questioned said that they watched less T.V. than they did a year ago, 22% said they watch the same amount, and 8% said they watch more. Find the probability that at least one will say they watched less T.V. this year than last year.
P(at least one) could be computed by adding the probabilities for r = 1, 2, 3, 4, and 5.
However, since r =0 is the complement of “at least one”, it is easier to use the complement rule.
Compute P(at least one) by 1- P(0).P(0) can be found by using the binompdf distribution, and is .002.P(at least one) = 1 - .002 = .998.
Assignments:Classwork: Pages 215-216 #1-14
AllHomework: Pages 216-219 #16-
32 Evens
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-3
More Discrete Probability Distributions.I. The Geometric Distribution
We use the Geometric distribution when we want to know how many times we may have to try something to get a success.
A. A Geometric Distribution is a discrete probability distribution of a random
variable x that satisfies the following conditions:1. A trial is
repeated until a success occurs.
a) You will take the English SOL test until you pass it, no matter how
many tries it takes!!!2. The
repeated trials are independent of each other.3. The
probability of success, p, is constant for each trial (which is just another
way of saying that they are independent).B. The good news is that the calculator
will also find the probability of a geometric distribution.
1. 2nd VARS will again get us to the list of distributions.
geometpdf and geometcdf are the last two on the list.2. Enter
the probability of success, the number of trials you are interested in, and press Enter.
PROBABILITY AND STATISTICSCHAPTER 4 NOTES SECTION 4-3
More Discrete Probability Distributions.II. The Poisson Distribution
We use the Poisson distribution when we want to know the probability that a specific number of occurrences takes place within a given unit of time or space.
(How likely is it that an employee will miss 15 days of work in a year?)
A. The Poisson Distribution is a discrete probability distribution of a random
variable x that satisfies the following conditions:1. The
experiment consists of counting the number of times x, and event, occurs in a given interval.2. The
probability of the event occurring is the same for each interval.3. The
number of occurrences in one interval is independent of the
number of occurrences in other intervals.B. The Poisson distributions are also on
the calculator, between the binom and geomet distributions.
1. Enter the average number of occurrences for the desired interval, and
the number you are interested in.
Paste and Enter to find the probability.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESSECTION 4-3 EXAMPLES
You are a telemarketer. From experience, you know that the probability that you will make a sale on any given call is 0.23.
Find the probability that your first sale on any given day will occur on the fourth or fifth call.
A. This is a geometric distribution question, since you are being asked how
many times you will have to call to make a sale.1. We need
to add the probability of the first sale happening on the fourth
call to the probability of the first sale occurring on the fifth call.(This is an OR
question, so we add).2. 2nd
VARS geometpdf, .23, 4 gives us .105, and 2nd VARS geometpdf, .23, 5
gives us .081. Add
those up to get .186.
a) We have an 18.6% chance of the first sale occurring on the
fourth or fifth call.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESSECTION 4-3 EXAMPLES
You are a telemarketer. From experience, you know that the probability that you will make a sale on any given call is 0.23.
Find the probability that the first sale will occur before your fourth sales call.A) This is still a geometric distribution question,
since you are being asked how many times you will have to call to make a sale.
1. We need to add the probability of the first sale happening on the first
call to the probability of the first sale occurring on the second call to
the probability of the first sale happening on the third call.
2. We can use the geometcdf option here, using 3 as the upper limit.
2nd VARS geometcdf, .23, 3 gives us .543.
a) We have a 54.3% chance of the first sale occurring before the
fourth call.
PROBABILITY AND STATISTICSCHAPTER 4 NOTESSECTION 4-3 EXAMPLES
Two thousand brown trout are introduced into a small lake. The lake has a volume of 20,000 cubic meters.
What is the probability that three brown trout are found in any given cubic meter of the lake?
A) This is a Poisson distribution, since we are interested in how many brown
trout occur within a specified area.1. To find
how many trout would be expected to be found in each cubic meter of
water, divide the 2,000 fish by the 20,000 cubic meters of water in
the lake.
This gives us an expected value of 0.1 trout per cubic meter.2. On the
TI-84, go to 2nd VARS C (poissonpdf). We use the pdf function because
we are interested in PRECISELY 3 brown trout.Enter
0.1 (what we expect to find) for the first value, and 3 (what we want to
find) for the second value. 3. We have
about a 0.00015 probability of finding 3 brown trout in any
randomly selected cubic meter of the lake.
Assignments:Classwork: Page 226
#1-10 AllHomework: Pages
226-227 #11-24 All