Scribd
Upload
255 views

Correlation vs. Regression

This document discusses correlation and regression analysis techniques. It defines correlation as measuring the strength of the linear relationship between two variables, while noting that correlation does not imply causation. Two main types of correlation are discussed: Pearson's correlation coefficient for continuous data and Spearman's rank correlation for non-continuous or ranked data. Regression analysis is introduced as concerned with predicting the relationships between variables. Time series analysis is also briefly mentioned as measuring trends and seasonal variations over time for forecasting purposes.

Uploaded by

bhupender1989
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
255 views

Correlation vs. Regression

This document discusses correlation and regression analysis techniques. It defines correlation as measuring the strength of the linear relationship between two variables, while noting that correlation does not imply causation. Two main types of correlation are discussed: Pearson's correlation coefficient for continuous data and Spearman's rank correlation for non-continuous or ranked data. Regression analysis is introduced as concerned with predicting the relationships between variables. Time series analysis is also briefly mentioned as measuring trends and seasonal variations over time for forecasting purposes.

Uploaded by

bhupender1989
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 15

Correlation analysis-meaning & types of correlation, Karl

Person’s coefficient of correlation and spearman’s rank


correlation; regression analysis- meaning and two lines of
regression; relationship between correlation and regression co-
efficient. Times series analysis- measurement of trend and
seasonal variations; time series and forecasting.

Unit 2

A S Raheja
[email protected]
9717871155
Correlation vs. Regression
 Examples
 Price and supply
 Income & Expenditure
 Correlation analysis is used to measure strength of the
association (linear relationship) between two variables
 Correlation is only concerned with strength of the relationship.
 No causal effect is implied with correlation.
 Most widely used and widely abused statistical measure.
 Reduces range of uncertainty.
Types of Correlations

 Issues with Correlation


 Chance Coincidence
 Influence of one on other or mutual dependence
 Both being influenced by third variable
 Methods
 Scatter diagram
 Karl Pearson’s Coefficient of Correlation
 Rank Method
Scatter Diagram
+ Correlation, r>0 Non Linear correlation

Y Y

X X
- Correlation, r<0 No Correlation, r = 0

Y Y

X X
Type of Correlation
 A measure of the linear association between variables
 – Positive Correlation indicates positive linear relationship
 – Negative Correlation indicates a negative linear relationship
 – Values close to zero indicates no linear relationship
 It not affected by the units of measurement for x and y
variables
 – Pearson product moment correlation coefficient or Sample
correlation coefficient, r (used in case data is continuous)
 It is a numerical index that reflects the linear relationship between two
variables
 The values of the descriptive statistic range between a value -1 (perfect –ve
correlation) to +1(perfect positive correlation) , it is also referred to as Bi
variate
Correlation Formula 1

 xy
r 

xy
x y  XY
1 _ _
  ( X  X )(Y  Y )
n
1 2 1 2

 x n (X  X )
  y  n  (Y Y )
Where
r = sample correlation coefficient,
σxy = sample covariance
σx = sample standard deviation of x
σy = sample standard deviation of y
Calculating from Covariance
Knowing the covariance and the standard deviations of
each variable we can compute the sample correlation
coefficient, r

Covariance = 11,
σx = 1.49, σy = 7.93

So Pearson r = 11/(1.49 x 7.93) = 0.93


Correlation Formula 2

1
 XY  n
 X . Y
r
1 1
 X 
2

n
(  X ) 2
 n
Y 2
( Y ) 2
Solution
Assumed mean

1
 xy  n
 x. y
r
1 1
 n
x 
2
( x ) 2
 n
y 2
( y ) 2

x  X a
y Y b
Spearman’s Rank Correlation

 On some occasion it is not possible to measure the variables


quantitatively or exact magnitude of the variable can not be
determined
 Numerical value of beauty
 Measure taste of wine

6 D
2

rs  1 
N ( N 2  1)
Rules

 When Rank are given – above formula is sufficient


 When Ranks are not given – ranks are assigned by taking either the
highest value as 1 or the lowest value as 1
 Equal Ranks- in some cases, rank may be equal , in such case each
individual is given an average rank
Exercise 1
Exercise 2
Exercise 3

You might also like