Scribd
Upload
507 views20 pages

CCP403

This document summarizes a lecture on correlation analysis techniques including Spearman's rank correlation and probable error. It defines Spearman's rank correlation as a non-parametric measure of statistical dependence between two variables that is used when quantitative measures cannot be assigned ranks. It provides the formula and steps to calculate Spearman's rank correlation coefficient from ranked or unranked data. Probable error is defined as a measure of the reliability of the correlation coefficient, and its formula is provided. Examples are given to demonstrate calculating Spearman's rank correlation and probable error.

Uploaded by

api-3849444
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
507 views20 pages

CCP403

This document summarizes a lecture on correlation analysis techniques including Spearman's rank correlation and probable error. It defines Spearman's rank correlation as a non-parametric measure of statistical dependence between two variables that is used when quantitative measures cannot be assigned ranks. It provides the formula and steps to calculate Spearman's rank correlation coefficient from ranked or unranked data. Probable error is defined as a measure of the reliability of the correlation coefficient, and its formula is provided. Examples are given to demonstrate calculating Spearman's rank correlation and probable error.

Uploaded by

api-3849444
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 20

DEPARTMENT OF TECHNICAL EDUCATION

ANDHRA PRADESH
Name : B.VIJAYA KUMAR
Designation : Lecturer
Branch : Commercial & Computer
Practice (CCP)
Institute : GPW/Guntur
Year/Sem : II Yr Sem 4
Subject : QUANTITATIVE
TECHNIQUES-II
Sub-Code : CCP-403
Topic : CORRELATION ANALYSIS
Duration : 50 minutes
Sub Topic : spearman’s rank correlation
and probable Error
Teaching Aids : PPT, ANIMATION
CCP403.5 1
Objectives

On completion of this period, you would be


able to:

 Know the methods of calculation of correlation


by Spearman
 Know the Probable Error

CCP403.5 2
Recap

In the last class, YOU learnt :

 Karl Pearson’s Correlation Coefficient

CCP403.5 3
Spearman’s Rank Correlation
 Charles Edward Spearman developed this
method in 1904.

 useful when quantitative measure cannot be fixed

Ex: beauty, intelligence etc.,

 Denoted by rs

 rs ranges between +1 and -1

CCP403.5 4
Calculation of Rank Correlation
 Formula:
6 D 2
rs  1 
N3  N

 r s= Spearman’s Rank Correlation


 D = difference of ranks
 N = number of observations

CCP403.5 5
Rank correlation can be done in two
ways:

 When ranks are given.

 When ranks are not given.

CCP403.5 6
Steps when ranks are given

 Take the difference of the two ranks ie.,


(R1-R2) and denote these differences by
D.
 Square these differences and obtain the
total  D 2
 Apply the formula

CCP403.5 7
Steps when ranks are not given:

 Assign the ranks based on the highest (or)


lowest value for each variable i.e. R1 and R2
 Calculate the difference between the ranks i.e.
(R1 - R2) and denote by D
 Square the difference and obtain the total  D 2

 Apply the formula

CCP403.5 8
Example 1

Compute Rank Correlation for the data below:

X: 415 434 420 430 424 428


Y: 330 332 328 331 327 325

CCP403.5 9
Solution to Ex. 1
Calculation of Spearman’s Rank Correlation
X R1 Y R2 D = (R1-R2) D2
415 6 330 3 3 9
434 1 332 1 0 0
420 5 328 4 1 1
430 2 331 2 0 0
424 4 327 5 -1 1
428 3 325 6 -3 9
∑D2 = 20

r = 1 - 6∑D2
N3-N
= 1- 6(20) =0.429
63-6 CCP403.5 10
Probable Error

 Helps in interpreting COC value


 Determines reliability of COC
 Formula : PE = 0.6745 1-r2
√N

Where r = Coefficient of Correlation


N = No of pairs of observations

CCP403.5 11
PE - Application

 If ‘r’ < PE, Correlation not significant


 If ‘r’ is 6 times greater than PE,
Correlation is significant
 Formula to get the lower and upper limits in
which COC lies
p = r + PE
‘p’ is called rho

CCP403.5 12
Probable Error - Conditions

 Data must form bell - shaped curve


 PE should be calculated from a sample
 No bias in selecting sample

CCP403.5 13
Example 2

If r = 0.6 and N=64, find out PE

Sol:

PE = 0.6745
1 r 2

N
r = 0.6 and N = 64
PE = 0.6745 1  (0.6) 2

64
= 0.054

CCP403.5 14
Summary:

 Spearman’s Rank Correlation is calculated


when quantitative measures are not suitable
for certain factors
Eg: Beauty, leader ship, intelligence,
 Probable Error measures the reliability of
COC value

CCP403.5 15
Quiz

1) PE is calculated when:
(a) Data looks like a bell-shaped curve
(b) Calculation of regression not possible
(c) Data is outside the sample
(d) None of the above

CCP403.5 16
Quiz

1. Spearman’s Rank Correlation is used:


(a) To calculate PE
(b) To measure beauty, leadership etc.,
(c) To draw a pie diagram
(d) All of the above

CCP403.5 17
Frequently Asked Questions:

1. Find the value of ‘n’ if the value of r is 0.64 and


its probable error (PE) is 0.1312
(march- 2008, C-05-CCP-403)

CCP403.5 18
Frequently Asked Questions:

2. Calculate spearman's coefficient of correlation


between marks assigned to ten students by judges
X and Y in a music completion is shown below:

(X) 52 53 42 60 45 41 37 38 25 27
(Y) 65 68 43 38 77 48 35 30 25 50

(marks 2008,C-05,CCP-403)

CCP403.5 19
Assignment:

2) What is a Probable Error?


3) Write the features of PE?
4) Explain briefly features of Spearman’s
Rank Correlation

CCP403.5 20

You might also like