Scribd
Upload
32 views22 pages

Regression Analysis

Uploaded by

2002rohanjha
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
32 views22 pages

Regression Analysis

Uploaded by

2002rohanjha
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 22

BUSINESS STATISTICS:

Text and Problems


With Introduction to Business Analytics

Dr. N. D. VOHRA
Chapter 13

regression analysis
REGRESSION

• Regression deals with the problem of prediction or estimation of


the value of a variable from the known value of another variable
or variables to which it is related.
• Regression analysis is the statistical tool which is employed for the
purposes of making estimates.
• Regression is a statistical model – a set of mathematical formulas
and assumptions, which are used to describe a real world situation.
• Although we can attempt the model to explain as much as
possible, it will not explain everything, and thus we will always
have some errors.
REGRESSION ANALYSIS:
AN INTRODUCTION

• Regression analysis is used to:


❑ Predict the value of a dependent variable based on the
value of at least one independent variable
❑ Explain the impact of changes in an independent
variable on the dependent variable
• Dependent variable: It is the variable we wish to explain
• Independent variable: This is the variable used to explain
the dependent variable
ESTIMATION AND REGRESSION

• Prediction:
Social • For social variables like unemployment level, criminal activities
Sciences etc.

• Prediction:
Natural • For medical sciences, problems related to atmosphere, etc.
Sciences

• Estimation :
Business • Production targets, Sales, Human Resources
Problems
SIMPLE AND MULTIPLE REGRESSION

Simple Regression Multiple Regression


•Two variables involved: • Variables involved: One
one dependent variable dependent and more than
and other independent one independent variable
variable •It is assumed that the
•It is assumed that the relationship is straight line
relationship is straight line type
type
ASSUMPTIONS OF LINEAR MODEL

• There is a linear relationship between the X and Y


variables. Like any straight line, it has two coefficients:
the intercept and the slope.
• The values of independent variable X are fixed while
those of dependent variable, Y, are random – with
randomness arising from the error term.
• Thus, for each value of X, there is a group of Y values.
ASSUMPTIONS OF LINEAR MODEL

• For each value of X, the set of Y values is normally distributed.


The means of these distributions of Y values all lie on the
regression line.
• All the distributions of Y values have equal variances.
• The Y values are all statistically independent so that in selection
of a sample, the Y values chosen for a particular value of X do
not depend on the Y values for any other X value.
ASSUMPTIONS OF MODEL

• There exists a linear relationship between X and Y


variables.
• The values of independent variable Y are fixed while
those of dependent variable X are random – with
randomness arising from the error term.
• The errors, є, are normally distributed with mean equal
to zero, and constant variance σ2.
• Further, they are independent in different observations.
PAIR OF REGRESSION LINES

• Using a given set of data, two regression lines can be


obtained: one showing regression of Y on X and the
second one showing regression of X on Y.
• The two lines have different equations because one of
these is obtained by minimizing the squared vertical
deviations and the other by horizontal deviations.
• To differentiate we denote the slope coefficients as byx
and bxy .
PAIR OF REGRESSION LINES

• Either both the regression lines be positively sloped or both would


be negatively sloped.
• Each regression line passes through the mean values of X and Y.
• Therefore, the point of intersection of the two regression lines yields
mean values of X and Y.
• Closer the two lines, stronger is the correlation between the two
variables.
• The two regression lines overlap when the correlation is of perfect
order.
PAIR OF REGRESSION LINES
AN ALTERNATE APPROACH TO
OBTAIN REGRESSION EQUATIONS

•
AN ALTERNATE APPROACH TO
OBTAIN REGRESSION EQUATIONS

•
OBTAINING REGRESSION COEFFICIENTS

•
OBTAINING REGRESSION
COEFFICIENTS
• OPTION 3:
• Deviations taken from assumed means Ax and Ay:
OBTAINING REGRESSION
COEFFICIENTS
• OPTION 4:
• Without measuring deviations
RELATIONSHIP BETWEEN REGRESSION
AND CORRELATION COEFFICIENTS

•
RELATIONSHIP BETWEEN REGRESSION
AND CORRELATION COEFFICIENTS

•
PROPERTIES OF
REGRESSION COEFFICIENTS

• While the coefficient of correlation cannot exceed ±1, there


are no such limits for regression coefficients.
• However, both the regression coefficients cannot be greater
than 1 (ignoring sign for negative coefficients)
• The regression coefficients are independent of the change
of origin but they are not independent of the change of
scale
• The two regression coefficients are reciprocal of each other
in case of perfect correlation between the variables.
CORRELATION Versus
REGRESSION ANALYSIS

CORRELATION REGRESSION

Discovers Describes
extent of Nature of
relationship relationship

Prediction/
Estimation of
No prediction
dependent
variable
END OF CHAPTER 13

You might also like