Class 1

Decision 516: Quantitative Business Analysis

 Statistics in Business Decisions

 Introduction to probability
 Conditional Probability
 Independence

Statistics and Walmart’s Brand Satisfaction

 Walmart’s brand satisfaction
Walmart is considering a new marketing campaign in Maryland. Before Walmart moves forward,
the head of the Marketing Department is concerned that Walmart’s brand is already well
perceived in that state compared to other states.

 A recent focus group study interviewed 2000 consumers in Maryland and 3500 consumers in
Mississippi. It concludes that 8% of the consumers are satisfied with Walmart’s brand in
Maryland while 12% are satisfied with Walmart’s brand in Mississippi.

Based on this information, what advice would you give?

 Walmart’s brand satisfaction
Walmart is considering to start a new marketing campaign in Maryland. Before Walmart moves
forward, the head of the Marketing Department is concerned that Walmart’s brand is already well
perceived in that state compared to other states.

 A recent focus group study interviewed 2000 consumers in Maryland and 3500 consumers in
Mississippi. It concludes that 8% of the consumers are satisfied with Walmart’s brand in
Maryland while 12% are satisfied with Walmart’s brand in Mississippi.

 She also considered the household income: below 50K (low), or above 50K (high). In
Maryland, among the 1300 high income households 31 were satisfied with Walmart’s brand
while among low income households 129 were satisfied. In Mississippi, 26 of the 1295 high
income households were satisfied in contrast to 394 satisfied low income households .

Based on this additional information, what advice would you give?

 Class 1
Decision 516: Quantitative Business Analysis

Topics:
 Introduction to probability (vocabulary and definitions)
 Conditional Probability
 Independence

Roadmap: where are we?

REGRESSION
ANALYSIS

ESTIMATION
CORRELATION
SAMPLING & CLT

NORMAL DISTRIBUTION

RANDOM VARIABLES

PROBABILITY THEORY
 Probabilities
Match.com ad on FB is viewed by a 28y/o man
price of AMZN at closing tomorrow up 2%
Blue Devils in Final Four this season
United States President is re-elected

 measure of uncertainty
 Where do the come from? Historical data, expert opinions, subjective

 quantifying uncertainty
 crucial in many industries (insurance, finance, marketing, energy, etc…)
 need for precision and clarity

Why quantifying uncertainty matters?

 What probability would you assign to the following verbal statements?

 “Usually”

 “Possible”

 “Somewhat likely”

 “Probably”
 Sample Space and Events
• Sample space: all possible outcomes of the experiment
• denoted as S
• Examples:
• die roll:
• ads on Facebook:

• Empty Set: denoted as ∅ (no outcomes)

• Event: A ⊆ S is a subset of the sample space
• Examples:
• die roll:
• ads on Facebook:

• Complement: A = “A does not occur” (also written as 𝐴̅ or ~𝐴)

• Examples:
• die roll
• ads on Facebook

Probability

 The probability of each event is a number between zero and one:

 𝟎 𝐏 𝐀 𝟏

 P S 1 (something occurs)
 P A 1 P A
 P ∅ 0

 Examples
 die roll:

 ads on Facebook:
 Combining events
 Intersection of some events means that all events have happened
 The intersection of two events A and B is denoted A ∩ B
 A ∩ B means both A and B occur
 In general, there is no simple formula for P A ∩ B
 Union of some events means that any of the events may have happened
 The union of two events A and B is denoted A ∪ B
 A ∪ B means that A or B occurs
 Probability of the union formula: P A ∪ B P A P B P A∩B
 The events A and B are mutually exclusive if A ∩ B ∅
 In this case, P A ∩ B 0;
 so when events A and B are mutually exclusive P A ∪ B P A P B
 A collection of event is collectively exhaustive if their union equals the sample space;
 A, B, C are collectively exhaustive if A ∪ B ∪ C S
 Probability tables
• Useful tool for organizing info about probabilities
Columns are labeled by
mutually exclusive and This entry is
B B always 1
collectively exhaustive
events A P A∩B P A∩B P A because it
corresponds to
P S
Rows are labeled by A P A ∩B P A ∩B P A
mutually exclusive and
collectively exhaustive P B P B 1
events

• The entries in the interior cells correspond to probabilities of the intersection events.
• The marginal entries correspond to probabilities of the marginal events and are always equal to
the sum of the interior entries in the corresponding row or column.
 Conditional probabilities
 The main idea is that probability valuations might change when additional information
becomes available.

 Example (die roll):

 A={even roll} ={2,4,6}, B={high roll}={4,5,6}
 P(A)= ½
 Suppose you know that the roll is high (B occurred).
 Does this have any effects on your valuation of the probability of an even roll?
 We need new notation to express this idea.

Notation:
P A|B probability of A given that B occurred

Conditional probabilities
P A|B = probability of A given B

S
B
B

A AB

Conditional Probability Formula:

P A∩B
P A|B
P B
 (Back to) Walmart ads on FB

 Walmart contracts 1 billion of additional impressions. Provide a simple forecast for the
number of clicks generated by them.

 If Walmart’s campaign were to target only women above 24, would you revise the above
forecast?

 Based on these calculations, would you recommend Walmart to implement such a targeted
campaign instead?
 Independence
Two events are independent (statistically) if the information about one of them happening, or
not happening, does not change the probability of the other one.
• In other words, A and B are independent if P(A|B)=P(A)
• e.g., with die roll, are A={1,3,5} and B={3,6} independent?

• There are three ways recognize independent events:

• A and B are independent if any of the following holds true:
• P A∣B P A
• P B∣A P B
• P A∩B P A P B
If any of the above equalities is true, then all three are true (can you show this with math?)
Econ 160 - Intro to Econometrics Class 3

Independence
Two events are independent (statistically) if the information about one of them occurring, or
not occurring, does not change the probability of the other one.
• In other words, A and B are independent if P(A|B)=P(A)

• Independence is not necessarily intuitive, and it is a very hard concept to fully grasp (one
of the hardest in this course).
• It captures the idea that statistical independence is about learning from information.
• When events are not independent, learning one of them has happened changes one’s
assessment of the likelihood that the other may happen.
• Independence is about processing new information, and how that information may or may
not change what one thinks.
Econ 160 - Intro to Econometrics Class 3
 Probability trees

• This is another
useful tool for
organizing info
about
probabilities
• Trees are useful
when some of the
given information
is about
conditional
probabilities

How to “grow” your probability tree

B  Start by branching over a set of mutually
P B exclusive and collectively exhaustive
events.
 Each initial branch corresponds to one
such event (there could be more than two
branches in total)
 A good rule of thumb:You should know
the probabilities of these events
B (otherwise, the probability tree might not
help you much)
P B

Continue by branching each of the formed branches…

A probability tree
A
B P A|B P A∩B

P B Ac
P Ac ∩ B
P Ac|B

A P A ∩ Bc
Bc P A|Bc
P Bc
Ac
P Ac|Bc P Ac ∩ Bc
 Problem Set 1, question 9
(Machine adjustment in a manufacturing process)
In the setup of a manufacturing process, a machine is either correctly or incorrectly adjusted. The
probability of a correct adjustment is .90. When correctly adjusted, the machine operates with a
5% defective rate. However, if it is incorrectly adjusted, it has a 75% defective rate. In either
case, defects occur independently of each other.
 After the machine starts a production run, what is the probability that a defect is observed
when one part is tested?

 Suppose that the one part selected by an inspector is found to be defective. What is the
probability that the machine is incorrectly adjusted?
 Class 1: news you can use
 Fundamental probability facts: for any event A,
0 P A 1
P A 1 P A
P A∪B P A P B P A∩B
 Conditional probability formula:
P A∩B
P AB
P B
 Two events A and B are independent if any of the following holds:
P AB P A P BA P B P A∩B P A P B

 Problem solving:
 Probability tables (intersection information)
 Probability trees (conditional information)

Next time

 HIV Testing reading discussion

 prepare the reading and be ready to discuss it

 Random variables