Transcript Multiple Regression

DSCI 5180: Introduction to the Business
Decision Process
Spring 2013 – Dr. Nick Evangelopoulos
Lecture 1:
Introduction to 5180 class
Review of Basic Statistics (Ch. 1-2)
slide 1
DSCI 5180
Decision Making
Business Decision Making
“Decisions, decisions, decisions…
Eeny, meeny, miny, moe!”
Miss Piggy, The Muppets (2011)
http://www.dilbert.com/2001-08-12/
slide 2
DSCI 5180
Decision Making
DSCI 5180 Learning Goals
All business decisions require valid data and valid analytical
techniques. The goals of this course include:
G1. Developing an appreciation for the role of statistics in
making decisions,
G2. Reviewing the central concepts of statistical analysis,
G3. Understanding Simple Regression/ Correlation as a data
analysis technique,
G4. Building models using Multiple Regression,
G5. Understanding the role of ANOVA in experimental designs,
and finally
G6. Developing the capability to analyze data to enable better
business decisions.
slide 3
Chapter 1
An Introduction to Regression
Analysis
Terry Dielman
Applied Regression Analysis:
A Second Course in Business and
Economic Statistics, fourth edition
Introduction
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc.
4
A Mountain of Data
Advances in technology have buried
present-day managers under a mountain
of data.
 This text has been prepared to give future
managers some tools for examining
relationships between two or more
variables.
 Some examples are how sales are affected
by advertising, or what determines the
selling price of a house.

Introduction
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc.
5
Regression Analysis
One of the most important tools for
examining relationships between
variables.
 You develop an equation for predicting a
dependent variable from one or more
explanatory variables.
 In the process, you also describe how the
relationship operates and sometimes how
to control the dependent variable.

Introduction
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc.
6
Trial and Error
 Much
statistical analysis is a
multistage process of trial and error.
 There is a good deal of exploratory
work, then several stages of model
building and judgment.
 The emphasis of this text is on the
process rather than computations or
theory.
Introduction
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc.
7
Software
Three software packages are discussed in
the text.
 The first is Excel because it is so often
used in business.
 Minitab is an efficient standalone package
that has been around since the 1970s.
 SAS is an all-encompassing package that
does many things other than statistical
analysis.

Introduction
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc.
8
Data, Data, Data
 Data
sets for all the examples and
exercises are on the CD.
 They come in versions for all three
packages.
 Each chapter ends with a section
illustrating how to apply the
techniques with the software.
 On these PowerPoint slides, almost
all of the output is from Minitab.
Introduction
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc.
9
DSCI 5180
Decision Making
HW 1
Warranty Calculation
The average lifetime of tires produced at a tire factory is mu=50K
miles. The standard deviation is 5K miles. If a 40K milewarranty is offered to the consumers, what is the expected
proportion of tires covered under the warranty?
NOTE: This is suggested homework, aimed at helping you
understand the course concepts and prepare for exams. DO
NOT TURN IN.
slide 10
DSCI 5180
Decision Making
The probability below 40
Z = ( X - )/
= (40 – 50)/5
= -10/5 = -2.00
.0228
-4
-3
-2
40
-1
0
1
2
3
4
50
P(Z < -2.00) = 0.0228
slide 11