Homework: Chapter 8: #1,14, 27 Random Variables

Due Wed, October 27 What we will cover this

week and early next week:
Announcements: Today: 8.1 to 8.3
• Quiz starts after class today, ends Wed. Chapter 8 Wed: 7.7 + extra material
• Order switched from original plan – start Friday: 8.4 + extra material,
Chapter 8 today, finish Chapter 7 on Wed. start 8.5 if time
Monday: 8.5 to 8.7
Skip Section 8.8

Copyright ©2004 Brooks/Cole, modified Oct 2010 by Jessica Utts

8.1 What is a Random Variable? Examples of Discrete Random Variables

Assigns a number to each outcome in the sample space for a
random circumstance, or to each unit in a population.
Random Variable: assigns a number to 1. Couple plans to have 3 children.
each outcome of a random circumstance, or,
The random circumstance includes the 3 births,
equivalently, to each unit in a population.
specifically the sexes of the 3 children.
Two different broad classes of random variables: Possible outcomes (sample space): BBB, BBG, etc.
X = number of girls
1. A continuous random variable can take any
X is discrete and can be 0, 1, 2, 3
value in an interval or collection of intervals.
For example, the number assigned to BBB is X=0
2. A discrete random variable can take one of a
countable list of distinct values. 2. Population consists of UCI students (unit = student)
Y = number of siblings a student has
Notation for either type: X, Y, Z, W, etc.
Y is discrete and can be 0, 1, 2, …??
Copyright ©2004 Brooks/Cole, modified Oct 2010 by Jessica Utts 3 4

Examples of Continuous Random Variables Today: Discrete Random Variables

Assigns a number to each outcome of a random X = the random variable (r.v.), such as number of girls.
circumstance, or to each unit in a population. k = a number the discrete r. v. could equal (0, 1, etc).
1. Population consists of UC female students P(X = k) is the probability that X equals k.
Unit = female student
W = height Probability distribution function (pdf) for a discrete r.v. X is a
W is continuous – can be anything in an interval, table or rule that assigns probabilities to possible values of X.
even if we report it to nearest inch or half inch Cumulative distribution function (cdf) is a rule or table that
provides P(X ≤ k) for every real number k. (More useful for
2. You are waiting at a bus stop for the next bus continuous random variables than for discrete, as we will see.)
Random circumstance = when the bus arrives
Y = time you have to wait NOTE: Sometimes the probabilities are given or observed, and
sometimes you have to compute them using rules from Ch. 7.
Y is continuous – anything in an interval
5 6
Conditions for Probabilities Another example of computing the PDF and
for Discrete Random Variables CDF from Chapter 7 Rules
Condition 1 Example: You buy 2 tickets for the Daily 3 lottery (different days)
The sum of the probabilities over all possible values Probability that you win each time is 1/1000 = .001, independent.
X = number of winning tickets you have, could be 0, 1, 2.
of a discrete random variable must equal 1.
P(X = 0) = (.999)2 = .998001 (Rule 3b) (998,001 in a million)
Condition 2 P(X = 2) = (.001)2 = .000001 (Rule 3b) (1 in a million)
The probability of any specific outcome for a P(X = 1) = 1 – P(X = 0 or X = 2)= 1 – (.998001 + .000001)
discrete random variable, P(X = k), must be between = .001998 (Rule 1) (1998 in a million)
0 and 1.
k pdf P(X=k) cdf P(X ≤ k)
Note: The possible values k are mutually exclusive
0 .998001 .998001
Example on Board: 2 Clicker questions with 4 choices each, 1 .001998 .999999
X = points earned if you are just guessing. Find pdf and cdf.
7 2 .000001 1.0 8

Example of using observed

Clicker data collection (non credit)
proportions to create a pdf
Survey of 173 students in introductory statistics: How many siblings (brothers and sisters) do
Number with k pdf cdf you have? Count half-siblings (share one
k
siblings P(X=k) P(X ≤ k) parent), but not step siblings.
0 14 14/173 = .08 .08
1 68 68/173 = .39 .39 + .08 = .47 A. 0
2 53 .31 .47 + .31 = .78 B. 1
3 21 .12 .90 C. 2
4 8 .05 .95 D. 3
5 6 .03 .98 E. 4 or more
6 3 .02 1.00 9

Graph of pdf for number of siblings (with More Complicated Examples for Discrete R.V.s
frequency instead of relative frequency)
Probability distribution function (pdf) X is a table or
Compare class results. rule that assigns probabilities to possible values of X.
Using the sample space to find probabilities:

Step 1: List all simple events in sample space.

Step 2: Find probability for each simple event.
Step 3: List possible values for random variable X
and identify the value for each simple event.
Step 4: Find all simple events for which X = k, for
each possible value k.
Step 5: P(X = k) is the sum of the probabilities for
all simple events for which X = k.
11 Copyright ©2004 Brooks/Cole, modified Oct 2010 by Jessica Utts 12
Example: Sibling blood types Example: Sibling blood types
Blood types and possible alleles: Family has 3 children. Probability of type A is ½ for each
Type O: Must be OO child. What are the probabilities of 0, 1, 2, or 3 with type A?
Type A: Could be AA or OA
Type B: Could be BB or OB Sample Space: For each child, write either O or A. There are eight
Type AB: Must be AB possible arrangements of O and A for three births. These are
the simple events.
Suppose father has OO (type O) and mother has OA (type A).
S = {OOO, OOA, OAO, AOO, OAA, AOA, AAO, AAA}
They have 3 children. Let X = number with Blood type A.
Sample Space and Probabilities: The eight simple events are
Each child equally likely to inherit: equally likely. Each has probability (1/2)(1/2)(1/2) = 1/8
Father Mother Child blood type Random Variable X: number of Type A in three children. For
O O Blood type O each simple event, the value of X is the number of A’s listed.
O A Blood type A
So, child has Type O or Type A, each with probability 1/2
13 14

How Many Children with Type A? Cumulative Distribution Function

Value of X for each simple event:
for number of Type A:
Simple Event OOO OOA OAO AOO OAA AOA AAO AAA
Probability 1/8 1/8 1/8 1/8 1/8 1/8 1/8 1/8 Cumulative distribution function (cdf) provides
X = # Type A 0 1 1 1 2 2 2 3 the probabilities P(X ≤ k) for any real number k.
Probability distribution function for X = # of Type A: Cumulative probability = probability that X is less
than or equal to a particular value.
Example: Cumulative Distribution Function
3/8
for the Number Kids with Type A

2/8
Probability

Graph of the pdf of X:

1/8

For example, the probability is 7/8 that ≤ 2 kids have Type A.

0
0 1 2 3
Number of Ty pe A 15 16

8.3 Expected Value (Mean) for

Example of expected value
Random Variables Number of siblings for intro stat students:
The expected value of a random variable is the
mean value of the variable X in the sample xi pi xi pi
space, or population, of possible outcomes.
0 14/173 = .08 .00 µ = E ( X ) = ∑ xi pi
1 68/173 = .39 .39
If X is a random variable with possible values x1, x2, x3, . . . , = 1.84
occurring with probabilities p1, p2, p3, . . . , 2 .31 .62
then the expected value of X is calculated as .12 .36 = mean number
3
of siblings
µ = E ( X ) = ∑ xi pi 4 .05 .20
5 .03 .15
6 .02 .12
Copyright ©2004 Brooks/Cole, modified Oct 2010 by Jessica Utts 17 18
Expected value = mean value is where the Other examples of expected value
picture of the pdf “balances”
Ex 1: Just guessing for 2 questions on clicker, quiz
X = clicker, Y = quiz; E(X)=2.5, E(Y)=1, (on board)
Ex 2: Raffle ticket costs $2.00. You win
$5.00 with probability 1/10, so net gain = $3
$100 with probability 1/100, so net gain = $98
Nothing with probability 89/100, so net gain = –$2
X = net gain. What is E(X)?
E(X) = $3 ×(10/100) + $98 ×(1/100) – $2.00 ×(89/100)
= $(30 + 98 – 178)/100 = –$50/100
This is a loss of 50 cents on average for each $2.00 ticket.
This means the people running the raffle gain 50 cents per
ticket.
1.84 19 20

Should you buy extended warranties? Notes about expected value

You buy a new appliance, computer, etc. • It’s the average or mean value of the random variable
over the long run.
• Extended warranty for a year costs $10.
• It may not be an actual possible value for the random
• Unknown to you, the probability you will need a
variable (usually it isn’t; e.g. 1.84 sibs).
repair is 1/50, and it will cost $200 if you do.
• In gambling, lotteries, insurance, extended warranty,
Is the warranty a good deal? etc., you can be pretty sure that your “expected” cost
X = your cost to repair the item. per event if you play or buy is more than if you don’t
If you buy the warranty – the house wins!
k P(X = k) k P(X=k)
$200 1/50 $200/50
your cost is fixed at $10. • However, for insurance, for example, you might
If you don’t, your cost is prefer the peace of mind of knowing your fixed cost.
$0 49/50 0/50
either $200 or $0, but the
For lottery, you might want the thrill of the possibility
E(X) = $200/50 = $4.00 long run average is $4.00
21
of winning, even though you lose on average. 22

Standard Deviation for a Example 8.13 Stability or Excitement

Discrete Random Variable Two plans for investing $100 – which would
you choose?
The standard deviation of a random variable is
essentially the average distance the random
variable falls from its mean over the long run.

If X is a random variable with possible values x1, x2, x3, . . . ,

occurring with probabilities p1, p2, p3, . . . , and expected
value E(X) = µ, then Expected Value for each plan:
Variance of X = V ( X ) = σ 2 = ∑ ( xi − µ ) pi
2
Plan 1:
E(X ) = $5,000×(.001) + $1,000×(.005) + $0×(.994) = $10.00
Standard Deviation of X = σ = ∑ (x − µ )
2
i pi
Plan 2:
E(Y ) = $20×(.3) + $10×(.2) + $4×(.5) = $10.00
Copyright ©2004 Brooks/Cole, modified Oct 2010 by Jessica Utts 23 Copyright ©2004 Brooks/Cole, modified Oct 2010 by Jessica Utts 24
Example 8.13 Stability or Excitement (cont) Notes about standard deviation
Variability for each plan: • Similar to when we used standard deviation for data
in Chapter 2, it is most useful for normal random
variables, which we will cover on Friday.
• In general, useful for comparing two random
variables to see which is more spread out. Examples:
Plan 1: Variance of X = $29,900.00 and σ = $172.92 – Two cities both have average yearly temperature of 65
Plan 2: Variance of X = $48.00 and σ = $6.93 degrees, but one has s.d of 5 degrees and the other has s.d.
of 20 degrees. Which would you prefer?
The possible outcomes for Plan 1 are much more variable.
– One investment fund has average rate of return over many
If you wanted to invest cautiously, you would choose
years of 8%, and s.d. of 2%. The other has average of
Plan 2, but if you wanted to have the chance to gain a
10%, but s.d. of 20%. The second one is higher on
large amount of money, you would choose Plan 1.
average, but is much more volatile.
Copyright ©2004 Brooks/Cole, modified Oct 2010 by Jessica Utts 25