Scribd
Upload
14 views

Linear Regression

The document discusses statistical methods focusing on linear regression and correlation analysis, explaining how regression can predict a dependent variable based on an independent variable, while correlation assesses the strength of the relationship between two variables. It includes examples of data analysis using the least-squares method and the coefficient of determination, along with practical exercises for estimating relationships in various contexts. The document emphasizes the distinction between regression and correlation, highlighting their respective purposes in statistical analysis.

Uploaded by

bilqeeshaider03
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
14 views

Linear Regression

The document discusses statistical methods focusing on linear regression and correlation analysis, explaining how regression can predict a dependent variable based on an independent variable, while correlation assesses the strength of the relationship between two variables. It includes examples of data analysis using the least-squares method and the coefficient of determination, along with practical exercises for estimating relationships in various contexts. The document emphasizes the distinction between regression and correlation, highlighting their respective purposes in statistical analysis.

Uploaded by

bilqeeshaider03
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

STATISTICAL

METHODS

Linear Regression and Correlation

DR. WAJID ALI SHAIKH


Dept. of Mathematics & Statistics
• Simple regression analysis provides an equation that can be used to
estimate or predict the value of a dependent variable for a given
value of an independent variable.
• In simple linear regression, the estimating equation has a graph
which is a straight line. The estimating equation is determined by
performing calculations on observed data.
• The purpose of simple linear correlation analysis is to determine
whether or not two variables are related, that is, whether one variable
tends to be larger/smaller as the other variable changes. The outcome
of correlation analysis is not an estimating equation but a conclusion
that there is, or is not, a relationship between the variables.
• Often, two variables are correlated, but an estimating (regression)
equation does a poor job of predicting the value of one variable from
the known value of the other variable. However, the correlation
between two variables aid in making a decision.
Cont…

• Summarized, we may say that ‘Regression Analysis’ is concerned


with measuring how one variable is related to another, that is, how
differences in one variable help to explain differences in other
variables.
• Correlation Analysis’ on the other hand, is simply concerned with
providing a statistical measure of the strength of any relationship
between variables.
• If this relationship only between two variables is being considered
the term simple regression and correlation analysis is used.
LEAST-SQUARES METHOD Cont…
THE COEFFICIENT OF DETERMINATION
o To understand the coefficient of determination, let us consider the example of
the amount of pollution at different distances in a river.
o We consider the point where the flow starts as zero point where no pollution
occurs.
o Then we measure the pollution at equal distances and the result is as follows.
Distance (x) 00 10 20 30 40 50 60
Amount of Pollution (y) 00 8 21 23 34 36 53
o The scatter diagram is also shown below along with the regression line.
Cont…
Cont…
Cont…
Example1: The following data show monthly expenditures in (Lacs of Rs.) and sales in (Millions of Rs.)
over a period of six months. Estimate the relationship between sales (y) and advertising
expenditures (x) by using least square regression line, and predict sales (y) for expenditure in
Solution:

7lacs of rupees (i.e., x=7). Also determine the coefficient of determination between the
expenditures and sales.

⟹ 𝒓 = 𝟎. 𝟒𝟒 = 𝟎. 𝟔𝟔
Example 2: Let us consider the following amount of pollution at different distances in a river and considered the point
where the flow starts as zero point and no pollution occurs. Using the data set, determine the predicted or
Solution:

trend line of linear regression also find the coefficient of determination.


Distance (x) 0 10 20 30 40 50 60
Amount of Pollution (y) 0 7 21 23 34 36 53
The necessary calculations are shown in the table below. It may be noted that we do not consider the first value
(0, 0) as it won’t disturb the entire computation. Thus we take n = 6

10 7 70 100 49
20 21 420 400 441
30 23 690 900 529
40 34 1360 1600 1156
50 36 1800 2500 1296
60 53 3180 3600 2809

Hence, this is the required linear regression line.


ASSIGNEMET-LINEAR REGRESSION AND CORRELATION
Q.1: An experiment was conducted on a new model of a particular make
of automobile to determine the stopping distance at various speeds. The
following data are recorded.
Speed 34 49 64 79 94 109
Stopping
15 25 40 61 87 118
distance
(a) Estimate the stopping distance, when the car is traveling at a speed of
70KM/h.
(b)Compute the coefficient of correlation and interpret your results.
Q.2: The following table shows the data on rainfall and discharge in a
certain river. Find the equation of the regression line to predict the
discharge from the rainfall. Estimate the discharge when the rainfall is
2.8. Also, compute the coefficient of determination
Rainfall (inches) 1.5 1.8 2.6 3.0 3.5
Discharge (1000 cc) 33 36 40 46 54
Q.3:The following data shows the cost of using CPU time versus time.
Find the regression line also Compute the coefficient of determination
and coefficient of correlation between time and cost.
CPU Time 5 8 15 20 23
Cost 125 175 350 460 520
Cont…

Q.4: The Voltage drop across a resistor for some different values of a
current, the results are
I
0.25 0.75 1.25 1.50 2.00
(Amperes)
V (Volts) 0.23 0.38 0.76 1.88 6.00
(a) Estimate the voltage drop for I = 0.9 and Interpret your result.
(b)Compute the coefficient of determination and coefficient of
correlation between voltage and current.
(c) Interpret your results.

Q.5: A sample of six days is produced in the following table showing the
relation between temperature and sale of deep freezers during a month.
Compute the regression line and coefficient of correlation. Estimate the
sale when the temperature reaches at 45oC.
Temperature oC 48 46 42 47 52 51
Sale of Deep-Freezers 26 28 20 25 28 29

You might also like