INBA 6152

Business Analytics and richard.coutou@fac.gsb.edu.tt

Quantitative Models
for Decision Making
mariawiltshire@fac.gsb.edu.tt
Business Analytics & Quantitative Analysis Models for
Decision Making
• This course aims to develop skills in the use of quantitative techniques commonly used
to provide insight into business and management decisions.

• An introduction to quantitative business techniques and business analytics and the ways
in which these techniques can be used to facilitate practical decision-making.

• Emphasis would be placed on formulation, model building, and interpretation of results

rather than theory.

• The course is decision and action oriented, not technique driven and numbers driven
hence the role of the computer and application software and the use of case studies will
be emphasized.
Business Analytics & Quantitative Analysis Models for
Decision Making
• Today's business problems tend to be very complex, and approaches such as business
experience, intuition, and thoughtful guesswork can no longer be applied to resolve
these situations.

• The quantitative techniques used in this course provide an objective basis to examine
business situations and make more informed judgments of the available alternatives.

• Topics in this course fall into two broad categories:

• Presentation, description, analysis, and interpretation of data.
• Decision-making with various quantitative tools.
Topics
1. Understanding Data

2. Statistical Measures

3. Visualizing Data

4. Probability Concepts and Applications

5. Decision Analysis and Simulation

6. Linear Regression

7. Forecasting

8. Linear Programming

9. Project Management

10. Transportation Models

Topics
1. Understanding Data
Presentation, description,
2. Statistical Measures analysis, and interpretation of
data.
3. Visualizing Data

4. Probability Concepts and Applications

5. Decision Analysis and Simulation

6. Linear Regression
Decision-making with
7. Forecasting various quantitative
tools.
8. Linear Programming

9. Project Management

10. Transportation Models

 Class Schedule
Session 1 Introduction to Quantitative Business Analytics
Monday 8th January, 2018 and Quantitative Models for Decision Making;
5:30pm-8:30pm Understanding Data
Richard Coutou
Maria Wiltshire
Session 2 Statistical Measures Supplement I
Monday 15th January, 2018
5:30pm-8:30pm
Maria Wiltshire
Session 3 Visual Analytics; STATA Supplement II
Monday 22nd January, 2018
5:30pm-8:30pm
Maria Wiltshire
 Class Schedule
Session 4 Probability Concepts and Normal Distribution Supplement III
Monday 29th January, 2018
5:30pm-8:30pm
Maria Wiltshire
Online Quiz Quiz I
Sunday 4th February, 2018
Session 5 Forecasting Individual Assignment (All
Monday 5th February, 2018 Students)
5:30pm-8:30pm
Maria Wiltshire
Session 6 Regression Models Group Assignment 1 - Food
Monday 19th February, 2018 Truck Forecaster
5:30pm-8:30pm
Maria Wiltshire
 Class Schedule
Session 7 Decision Analysis Group Assignment 2 - George’s
Monday 26th February, 2018 T-Shirts
5:30pm-8:30pm
Richard Coutou
Online Quiz Quiz II
Sunday 4th March, 2018
Session 8 Simulation Food Truck Forecaster
Monday 5th March, 2018 Presentations (no name group)
5:30pm-8:30pm Submission of Group
Richard Coutou Assignment I (All Groups)
Session 9 Linear Programming George’s T-Shirts Presentations
Monday 12th March, 2018 (Zenith, Stargirls 61)
5:30pm-8:30pm Submission of Group
Richard Coutou Assignment II (All Groups)
 Class Schedule
Session 10 Linear Programming Group Assignment 3
Monday 19th March, 2018 Transportation Models Merton Truck Company
5:30pm-8:30pm
Richard Coutou
Session 11 Project Management Individual Assignment
Monday 26th March, 2018 Submission (All Students)
5:30pm-8:30pm
Richard Coutou
Online Quiz Quiz III
Sunday 1st April, 2018
Session 12 Course Review Merton Truck Company
1 session – to be advised Presentations (Mental Models,
Richard Coutou Gates 61)
Maria Wiltshire Submission of Group
Assignment III (All Groups)
Text
• Title: Quantitative Analysis for Management
Author: Render, Stair, Hanna and Hale
ISBN-13 978-1-292-05932-7
Edition – 12th
Publisher: South Western
Text
• Title: Quantitative Analysis for Management
Author: Render, Stair, Hanna and Hale
ISBN-13 978-0-273-75286-8
Edition – 11th
Publisher: South Western
 Assessment
Method Percentage
Case Studies - Group 30%
(Presentation and Report)
(3 Cases; 10% each)
Individual Online Quizzes 15%
(3 online quizzes; 5% each)
Individual Assignment 10%
Class Contribution 5%
Final Examination 40%
TOTAL 100%
 Case Studies
• 3 Case Studies

• Each group must prepare each case study.

• Presentation and a short written report for submission.

• Presentations to be no more than 10 minutes.

• Written Report 10 - 12 pages.

Quizzes

• 3 Quizzes

• 15 to 20 minutes long.

Quiz Dates

Sunday 4th February, 2018

Sunday 4th March, 2018
Sunday 1st April 2018
Individual Assignment

• Assignment

• 10% of Final Grade

• Begins on Monday 5th February, 2018

• Submission Monday 26th March, 2018

Class Contribution

• 5% of final grade

• Students are expected to actively participate in class discussions and

presentations.

• Grades are determined by the course facilitator/s.

Final Examination
• 40% of final grade
• 3 Hour Examination
• Section 1 - Compulsory questions
• Section 2 – Choice (e.g. Section has 4 questions, choose any 2 questions)
• Allowed to use a Scientific Calculator
• Will be given Standard Normal Tables or Random Number Tables if
required.
Tutorials

3 Tutorials

• Dates to be determined by Facilitators with consultation from the

course representative.

• Also dependent on the availability of classrooms.

• Some topics will be covered, however, students should come

prepared with questions.
What is Quantitative Analysis?
Quantitative Business Analysis
• Scientific approach to managerial decision making.

• In solving a problem, managers and decision makers must consider

both quantitative and qualitative factors.

• Computers have been instrumental in the increasing use of

quantitative analysis.
 Quantitative Analysis Approach
1. Defining the problem
2. Developing the model
3. Acquiring input data
4. Developing a solution
5. Testing the solution
6. Analyzing the results
7. Implementing the results
1 - Defining the Problem
• Develop a clear, concise statement of the problem.

• Bad problem definition is a major reason for failure of management

science or operations research groups.
2 - Developing the Model
• Develop a model which is a mathematical representation of a
situation.

• The different types of models used are physical, scale, schematic, and
mathematical models.

• A mathematical model is a set of mathematical equations. Usually

expressed as equations or inequalities.

• Models contain variables and parameters.

2 - Developing the Model
• A variable is a measureable quantity that may vary or is subject to
change.

• A parameter is a measurable problem which is inherent to the

problem.

• All models should be developed carefully, solvable, realistic, easy to

understand and modify.
3 - Acquiring Input Data
• Data must be obtained once the model has been developed.

• It is essential to obtain accurate data.

• Garbage In, Garbage Out – Improper data will result in misleading

results.
4 - Developing a Solution
• This involves manipulating the model to arrive at the best/optimal
solution to the problem.

• In may entail solving an equation, the use of trial and error, trying
different approaches or picking a results with the best decision.

• Repeating particular steps to find the best solution is called an

algorithm.
5 - Testing the Solution
• A solution must be tested completely before it can be analysed.

• Testing the input data and the model includes determining the
accuracy and completeness of the data used by the model.

• Inaccurate data will lead to inaccurate solution.

6 - Analyzing the Results and Sensitivity
Analysis
• Analyzing the results starts with determining the implications of the
solution.

• A solution will result in some kind of action or change in a way an

organization is operating.

• The sensitivity of the solution to change in the model and input data
is a very important part of analyzing the results.
7 - Implementing the Results
• This is the process of incorporating the solution into the company or
organization.

• Implementing a solution may be difficult as managers and decision

makers may be resistant to change.

• After a solution is implemented, it must be closely monitored as there

may be numerous changes that call for modifications to the original
model or solution.
Advantages of Mathematical Modelling
1. Models can accurately represent reality.
2. Model can help a decision maker formulate problems.
3. Models can give us insight and information.
4. Model can save time and money in decision making and problem
solving.
5. A model might be the only way to solve some large and complex
problems in a timely fashion.
6. A model can be used to communicate problems and solutions to
others.
Possible Problems with the Quantitative
Analysis Approach
• Defining the problem
• Conflicting viewpoints
• Impact of other departments
• Beginning Assumptions
• Solution outdated

• Developing A Model
• Fitting the text book models
• Understanding the model
Possible Problems with the Quantitative
Analysis Approach
• Acquiring Input Data
• Use accounting data
• Validity of data

• Developing A Solution
• Hard to understand mathematics
• Only one answer is limiting
Business Analytics & Quantitative Analysis Models for Decision
Making
Basic Mathematical Concepts

• Basic Arithmetic
• Algebraic Equations
• Percentages
• Graphs
• Statistics and Probability
Business Analytics & Quantitative Analysis Models for Decision
Making

Basic Financial Concepts

• Revenue

• Fixed Cost

• Variable Cost

• Contribution Margin

• Profit
Develop a Model
INBA 6152
Business Analytics and richard.coutou@fac.gsb.edu.tt
Quantitative Models
for Decision Making
mariawiltshire@fac.gsb.edu.tt
STATISTICAL
MEASURES
What are Statistical Measures?

• measures of central tendency

• measures of central location
• classed as summary statistics
• The mean (often called the average) is most
likely the measure of central tendency that you
are most familiar with, but there are others.
Understanding Data
A better understanding can be derived from answers to
two questions:

●Where is the dataset centred?

●How dispersed is the dataset about its centre’?

Two Sets of Measures

●The first question is answered by way of measures of

central tendency

●The answer to the second question resides in measures

of dispersion
Measures of Central Tendency
●A measure of central tendency for a dataset is a number
that indicates the ‘centre’ of the distribution of the
dataset.

●Three main measures of central tendency:

1.Mean 2. Mode 3. Median
Measures of Dispersion
●Range
●The Interquartile Range
●Quartiles
●Variance
●Standard Deviation
• Once you’ve collected some data a very useful thing to do is
to plot a graph of how many times each score occurs.

• This is known as a frequency distribution, or histogram,

which is a graph plotting values of observations on the
horizontal axis, with a bar showing how many times each
value occurred in the data set.

• Frequency distributions can be very useful for assessing

properties of the distribution of scores.
The Centre of a Distribution
We can also calculate where the centre of a frequency
distribution lies (known as the central tendency).
There are three measures commonly used:
• the mean
• the mode
• the median
The Mode
• The mode is simply the score that occurs most frequently in the data
set.
• This is easy to spot in a frequency distribution because it will be the
tallest bar!
• To calculate the mode, simply place the data in ascending order (to
make life easier), count how many times each score occurs, and the
score that occurs the most is the mode!
The Mode
• One problem with the mode is that it can often take on
several values.
• For example, Figure 1 shows an example of a distribution
with two modes (there are two bars that are the highest),
which is said to be bimodal.
• It’s also possible to find data sets with more than two modes
(multimodal). Also, if the frequencies of certain scores are
very similar, then the mode can be influenced by only a small
number of cases.
Figure 1 : A Bimodal Distribution
 The Mode
● The Mode may not necessarily be affected by a change in
one data point

● It may not be unique

● If it is unique, the dataset is unimodal

● Otherwise the dataset may be bi-modal, multi-modal or

even possess no mode at all

● The sum of the deviations of the data points from the

Mode is not guaranteed to be zero
The Median
• Another way to quantify the centre of a distribution is to look for the
middle score when scores are ranked in order of magnitude. This is
called the median.
• For example, If I am a business owner, selling candles and I did some
research in a particular area and I obtained the prices of a particular
brand of candles at different stores.
• Imagine we looked at these prices from 11 different stores (which we
call scores for each store):-
108, 103, 252, 121, 93, 57, 40, 53, 22, 116, 98.
The Median
• To calculate the median, we first arrange these scores into ascending
order: 22, 40, 53, 57, 93, 98, 103, 108, 116, 121, 252.
• Next, we find the position of the middle score by counting the
number of scores we have collected (n), adding 1 to this value, and
then dividing by 2.
• With 11 scores, this gives us (n + 1)/2 = (11 + 1)/2 = 12/2 = 6. Then,
we find the score that is positioned at the location we have just
calculated. So, in this example we find the sixth score:98
The Median
• This works very nicely when we have an odd number of scores (as in
this example) but when we have an even number of scores there
won’t be a middle value.
• Let’s imagine that we decided that because the highest score was so
big (more than twice as large as the next biggest number), we would
ignore it.
• We have only 10 scores now. As before, we should rank-order these
scores: 22, 40, 53, 57, 93, 98, 103, 108, 116, 121. We then calculate
the position of the middle score, but this time it is (n + 1)/2 = 11/2 =
5.5
The Median
• This means that the median is halfway between the fifth and sixth
scores. To get the median we add these two scores and divide by 2
• In this example, the fifth score in the ordered list was 93 and the sixth
score was 98.
• We add these together (93 + 98 = 191) and then divide this value by
(191/2 = 95.5). The median price was, therefore, 95.5
The Median
• The median is relatively unaffected by extreme scores at
either end of the distribution: the median changed only from
98 to 95.5
• when we removed the extreme score of 252.
• The median is also relatively unaffected by skewed
distributions and can be used with ordinal, interval and ratio
data (it cannot, however, be used with nominal data because
these data have no numerical order).
 The Median
●It divides the dataset into halves

●It is unique

●A change in one data point may not impact the Median

●A change in one outlier should not impact the Median

●It is relatively easy to compute

●It is a suitable measure of central tendency when the dataset is skewed

 What does a Quantitative Analyst do?
• As quantitative analysts, we are interested in discovering
something about a ‘real-world’ phenomenon.

• These real-world phenomena can be anything from the

behaviour of interest rates in the economic market to the
behaviour of consumer demand for a particular good.
• Whatever the phenomenon we desire to explain, we
collect data from the real world to test our hypotheses
about the phenomenon.

• As mentioned before, testing these hypotheses involves

building statistical models of the phenomenon of interest

• the statistical model we build must represent the data

collected (the observed data) as closely as possible. The
degree to which a statistical model represents the data
collected is known as the fit of the model.
Populations and Samples
• As researchers, we are interested in finding results that apply to an
entire population of people or things.
• For example economists may want to build models that apply to all
salaries.
• A population can be very general (all human beings) or very narrow
(all students that hate Mathematics). Usually, scientists strive to infer
things about general populations rather than narrow ones.
Samples
• Rarely, if ever, have access to every member of a population.

• Psychologists cannot collect data from every human being

and lecturers cannot observe every student that hates
Mathematics.

• Therefore, we collect data from a small subset of the

population (known as a sample) and use these data to infer
things about the population as a whole.
Simple Statistical Models
• One of the simplest models used in statistics is the mean.
• The mean is a statistical model of the data because it is a
hypothetical value that doesn’t have to be a value that is
actually observed in the data.
The Mean
• The mean is the measure of central tendency that you are most likely
to have heard of because it is simply the average score and the media
are full of average scores.
• To calculate the mean we simply add up all of the scores and then
divide by the total number of scores we have
 The Mean
● We define the Mean as the arithmetic average of the dataset

●It is easy to compute

●It is denoted by (∑Xi / N) for a dataset of N data points.

●It is unique

●The computation uses all data points in the dataset

●It responds to changes in any of the data points

The Mean
●It is possible that a few extremely large or extremely small
values may be part of a dataset; the remainder of the dataset
being clustered in some defined range.

●The Mean in such cases could be misleading since these

outliers (as they are called) or untypical values could inflate or
deflate the value of the mean
 For example:
If we took five statistics lecturers and measured
the number of friends that they had, we might
find the following data:
1, 2, 3, 3 and 4.

Lecturer No. of Friends

Lecturer 1 1
Lecturer 2 2
Lecturer 3 3
Lecturer 4 3
Lecturer 5 4
The Mean
• If we take the mean number of friends, this can be calculated
by adding the values we obtained, and dividing by the
number of values measured:
(1 + 2 + 3 + 3 + 4)/5 = 2.6
• Now, we know that it is impossible to have 2.6 friends
• Hence the mean value is a hypothetical value.
• As such, the mean is a model created to summarize our
data.
Assessing the Fit of the Mean
• With any statistical model we have to assess the fit
• With most statistical models we can determine whether the
model is accurate by looking at how different our real data
are from the model that we have created.
• The easiest way to do this is to look at the difference
between the data we observed and the model fitted.
Graph 1: Showing the difference between the observed number of friends
that each statistics lecturer had, and the mean number of friends
Assessing the Fit of the Mean
• The figure shows the no. of friends that each statistics lecturer had,
and also the mean number that we calculated earlier on.
• The line representing the mean can be thought of as our model, and
the circles are the observed data.
• The diagram also has a series of vertical lines that connect each
observed value to the mean value.
• These lines represent the deviance between the observed data and
our model and can be thought of as the error in the model.
Assessing the Fit of the Mean
• We can calculate the magnitude of these deviances by simply
subtracting the mean value (x_bar) from each of the observed values
(x_i).

• For example, lecturer 1 had only 1 friend

• So the difference is x_1 - x_bar= 1 - 2.6 = -1.6.
• You might notice that the deviance is a negative number, and this
represents the fact that our model overestimates this lecturer’s
popularity
• It predicts that he will have 2.6 friends yet in reality he has
only 1 friend

• Now, how can we use these deviances to estimate the

accuracy of the model?

• One possibility is to add up the deviances (this would give us

an estimate of the total error).

• If we were to do this we would find that……

Table 1: Showing how Observed values
deviate from the Mean
Lecturer No. of Friends Deviances
(Number of friends-Mean number of friends)

Lecturer 1 1 1-2.6=-1.6
Lecturer 2 2 2-2.6=-0.6
Lecturer 3 3 3-2.6=+0.4
Lecturer 4 3 3-2.6=+0.4
Lecturer 5 4 4-2.6=+1.4
Total Error
• Total Error = Sum of Deviances
=(-1.6)+ (-0.6)+(+0.4)+ (+0.40)+(+1.4)
=0
• So, in effect the result tells us that there is no total error between our
model and the observed data.
• Hence the mean is a perfect representation of the data.
Do you think the statement is true or false?

• “Hence the mean is a perfect representation of the data”

• Since our total error is zero
Is this statement True or False?
Now, this clearly isn’t true:
Why?
Is this statement True or False?
Now, this clearly isn’t true:
Why?
• there were errors but some of them were positive
• some were negative
• they have simply cancelled each other out.
How do we deal with this situation?
• It is clear that we need to avoid the problem of which direction (+ or -
-)the error is in and one mathematical way to do this is to

• square each error i.e. multiply each error by itself.

• So, rather than calculating the sum of errors, we calculate the sum of
squared errors.
• When you multiply a number by itself it becomes positive.
Sum Squared of Errors
• In this example = Sum of Squared Errors
=(-1.6)^2+ (-0.6)^2+(+0.4)^2+ (+0.4^2)+(+1.4)^2
=2.56+0.36+1.6+1.6+1.96
=5.20
Sum of Squared Errors
• sum of squared errors (SSE) is a good measure of the
accuracy of our model.
• However, it is fairly obvious that the sum of squared errors
is dependent upon the amount of data that has been
collected – the more data points, the higher the SS.
• To overcome this problem we calculate the average error by
dividing the SS by the number of observations (N).
Sum of Squared Errors (SS)
• If we are interested only in the average error for the sample,
then we can divide by N alone.

• However, we are generally interested in using the error in the

sample to estimate the error in the population

• So we divide the SS by the number of observations minus 1

• This measure is known as the variance

Variance
• Variance is, therefore, the average error between the mean
and the observations made (and so is a measure of how well
the model fits the actual data).

• There is one problem with the variance as a measure: it gives

us a measure in units squared (because we squared each
error in the calculation).

• In our example we would have to say that the average error

in our data (the variance) was 1.3 friends squared. i.e.(
5.2/4=1.3)
Variance
• It makes little enough sense to talk about 1.3 friends, but it makes
even less to talk about friends squared!
• For this reason, we often take the square root of the variance (which
ensures that the measure of average error is in the same units as the
original measure).
• This measure is known as the standard deviation and is simply the
square root of the variance. In this example the standard deviation
is: =square root of 1.3= 1.14
Measures of Fit
• The sum of squares, variance and standard deviation are all,
therefore, measures of the ‘fit’ (i.e. how well the mean represents the
data).
• Small standard deviations (relative to the value of the mean itself)
indicate that data points are close to the mean.
• Large standard deviations (relative to the mean) indicates that the
data points are distant from the mean (i.e. the mean is not an
accurate representation of the data).
• A standard deviation of 0 would mean that all of the scores were the
same.
 More Rigorous Measures of
Dispersion
●The more rigorous measures are Variance &
Standard Deviation

●Variance equals the mean of the squares of

the deviations from the mean

●Variance = [ ∑ (xi – Mean) 2]/ N

 Variance
Variance = [ ∑fi (xi – Mean) 2]/ N

●Variance is never negative

●It attaches a greater penalty to greater deviations from the

mean (i.e. The equivalent of the square of the value)

●The greater the dispersion the greater the variance

●Unfortunately it is not expressed in the same unit of measure

as the data points
 Standard Deviation
●Getting around the shortcoming of Variance requires that we
find its square root

●The square root will have the same unit of measure as the data
in the dataset.

●The square root of variance is called standard deviation.

●It retains all the desirable qualities of variance

 If a dataset that has the same
mean then does it have the same
standard deviation?
Example:
• Overall ratings (on a 5-point scale) were collected for 2 lecturers on 5
different lectures.

• After each of five different lectures, both lecturers had an average

rating of 2.6 out of 5 across the lectures.
SMALL STANDARD DEVIATION

• The first lecturer had a standard deviation of 0.55 (relatively small

compared to the mean).
• It should be clear from the graph that ratings for this lecturer were
consistently close to the mean rating.
• There was a small fluctuation, but generally his lectures did not vary
in popularity.
• As such, the mean is an accurate representation of his ratings.
• The mean is a good fit of the data.
LARGE STANDARD DEVIATION
• The second lecturer, however, had a standard deviation of 1.82
(relatively high compared to the mean).
• The ratings for this lecturer are clearly more spread from the mean;
that is, for some lectures he received very high ratings, and for others
his ratings were appalling.
• Therefore, the mean is not such an accurate representation of his
performance because there was a lot of variability in the popularity of
his lectures.
• The mean is a poor fit of the data.
The illustration should hopefully make
clear why the standard deviation is a
measure of how well the mean
represents the data