0% found this document useful (0 votes)
24 views10 pages

Linear Regression

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
0% found this document useful (0 votes)
24 views10 pages

Linear Regression

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/ 10

J.K.SHAH CLASSES S.Y.J.C.

- STATISTICS

CHAPTER 3 LINEAR REGRESSION

Basic Concepts
1. The technique used for predicting the value of one variable for a given value of
the other variable is called regression.
2. Regression is a statistical tool for investigating the relationship between
variables.
3. Value being predicted is called the response or dependent variable.
4. Value used for predicting the response or dependent variables are called
predictors or independent variables.
5. A linear regression model consists of a linear equation with unknown
coefficient (called as parameters of linear regression model.)
6. Correlation analysis is used for measuring the strength or degree of the
relationship & between the independent & dependent variable.

Regression Coefficients of:


a) X on Y:
=

̅ ̅
=
̅

̅ ̅
̅ ̅
= ̅ ̅

b) Y on X:
=

̅ ̅
=

: 306 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
̅ ̅
̅ ̅
= ̅ ̅

=
Where
 Covariance of X & Y:
v ̅ ̅
̅ ̅

 Variance of X
̅
̅ Or

 Variance of Y
̅
̅ Or

 Standard Deviation √

 Regression lines / Regression Equations


(a)
𝑥 →? 𝑥→ p d tv l
𝑦 → 10 𝑦→I d p d tv l

𝑥 p d d 𝑦

̅ ̅
(b) 𝑥 → 12𝐷𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡
𝑥 → I 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒
d p d tv l
𝑦 → ? 𝑦→ p d t v l
𝑦 p d d 𝑥

̅ ̅

𝐷𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒
 Karl Person’s Correlation coefficient/ product moment correlation coeff.
Denoted by
“r”

1 ≤𝒓≤1

or
√ √

: 307 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS

Note:
If 𝑥𝑦 & 𝑥 both are positive then r must be positive and vice-versa

 Identification of Regression lines


Given two regression lines without stating which is y on x & which is x on y.
i) Based on Assumption:
Follow the steps:
1. Assume one regression equation as Y on X & other as X on Y

2. Write Y on X equation as

Where →
Write X on Y equation as

Where →

3. Compute:

4. Check: O ≤𝑟 ≤1

If satisfied then assumption are correct else reverse the assumption

ii) Slope method (Alternate Method)


Steps:
1. From 1st Eqn.

Find:

2. From 2nd Eqn.

Find:

3. Compare
| | | | or | | | |

4. Allocate
(Whichever is less)
(Whichever is high)

: 308 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
5. Identify Reg. lines
If byx get from then 1st Eqn. is y on x & if bxy get from then 2nd
Eqn.is x on y and vice-versa.

 Equation already mentioned that given eqn. is X on Y or Y on X.

 To find ̅ , ̅:
 Line of regression have a point of intersection ( ̅ , ̅)
 Do Simultaneous Eqn. and find value of & .
Value of → ̅
Value of →̅

 Shift of origin & change of scale


 Regression coeff. are independent of change of origin but not of scale.
 bxy & byx are not affected by change of origin but are affected by change
of scale.
This property is known as Invariance property
 Invariant property states that bxy & byx are invariant under change of
origin but are not invariant under change of scale.
Let &
(i) 𝑐 𝑑
𝑏𝑥𝑦 uv 𝑏𝑦𝑥 vu
𝑑 𝑐

(ii) If c & d are of same sign


Then, 𝑟𝑥𝑦 𝑟𝑢𝑣

If c & d are of opp. Sign


Then, 𝑟 𝑟
𝑥𝑦 𝑢𝑣

 Properties of Regression Coeff.


1) Correlation Coeff. is symmetric
i.e. = but regression Coeff. are not symmetric

: 309 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
2) When 0 both regression coefficients are 0
3) When 1 then,
a) Two regression lines become identical i.e. they coincide.
b) = & =
4) If 1 then 1

5) | | | |

6)

Sign analogy of &

+ + +
- - -

7) Case I Case II
= + ve 𝑥𝑦 = - ve
= - ve 𝑦𝑥 = + ve

In Both Case

"𝑟 will be imaginary”


Data is InConsistent

: 310 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS

OBJECTIVES
I) Choose the correct alternative.
1. Regression analysis is the theory of
a) Estimation b) Prediction
c) Both a and b d) Calculation

2. We can estimate the value of one variable with the help of other known
variable only if they are
a) Correlated b) Positively correlated
c) Negatively correlated d) Uncorrelated

3. There are ___________types of regression equations.


a) 4 b) 2 c) 3 d) 1

4. In the regression equation of Y on X


a) X is independent and Y is dependent.
b) Y is independent and X is dependent.
c) Both X and Y are independent.
d) Both X and Y are dependent.

5. In the regression equation of X on Y


a) X is independent and Y is dependent.
b) Y is independent and X is dependent.
c) Both X and Y are independent.
d) Both X and Y are dependent.

6. is___________
a) Regression coefficient of Y on X
b) Regression coefficient of X on Y
c) Correlation coefficient between X and Y
d) Covariance between X and Y

7. is___________
a) Regression coefficient of Y on X
b) Regression coefficient of X on Y
c) Correlation coefficient between X and Y
d) Covariance between X and Y
8. ‘r’ is ____________
a) Regression coefficient of Y on X
b) Regression coefficient of X on Y
c) Correlation coefficient between X and Y
d) Covariance between X and Y

: 311 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
9. . _____________2
a) b) c) d)

10. If 1 then is__________


a) 1 b) 1 c) 0 d) 0

11. | | ___________
a) | | b) 2| | c) r d) 2r

12. and are_________


a) Independent of change of origin and scale
b) Independent of change of origin but not of scale
c) Independent of change of scale but not of origin
d. Affected by change of origin and scale

13. If u and then __________


a) b)

c) d)

14. If u and then __________


a) b)

c) d)

15. Corr ( ___________


a) 0 (b) 1 (c) −1 (d) can't be found

16. Corr =
a) corr( ) (b) corr
c) corr (d) cov

17. ( )
a) c and d are opposite in sign
b) c and d are same in sign
c) a and b are opposite in sign
d) a and b are same in sign

: 312 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
18. Regression equation of X on Y is________
a) ̅ ̅
b) ̅ ̅
c) ̅ ̅
d) ̅ ̅

19. Regression equation of Y on X is________


a) ̅ ̅
b) ̅ ̅
c) ̅ ̅
d) ̅ ̅

20. __________
a) b)

c) d)

21. __________
a) b)

c) d)

22. __________
a) ̅ ̅

̅ ̅
b)

c) ̅̅

d) b and c both

23. If 0 and 0 then ‘r’ is________


a) 0 b) 0 c) 1 d)not found

24. If equations of regression lines are 3 2 − 26 = 0 and 31


0 then means of and are
a) (7,4) b) (4,7) c) (2,9) d) (−4,7)

: 313 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
II) Fill in the blanks
1. If 0 and 0 then ‘r’ is________
2. Regression equation of Y on X is________
3. Regression equation of X on Y is________
4. There are________ types of regression equations.
5. Corr = ________
6. If u and then __________
7. If u and then __________
8. | | ___________
9. If 1 then is__________
10. _________

III) State whether each of the following is True or False.


1. Corr (X, X) =1

2. Regression equation of X on Y is
̅ ̅

3. Regression equation of Y on X is
̅ ̅

4. Corr = Corr

5. and are independent of change of origin and scale.

6. ‘r’ is regression coefficient of Y on X

7. is correlation coefficient between X and Y

8. I d t

9. If and then =r

10. In the regression equation of Y on X, represents slope of the line.

: 314 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
Answers:
I) 1) c
2) a
3) b
4) a
5) b
6) b
7) a
8) c
9) c
10) b
11) b
12) b
13) a
14) b
15) b
16) c
17) a
18) b
19) a
20) b
21) a
22) d
23) b
24) b

II) 1) Negative
2) ̅ ̅
3) ̅ ̅
4) 2
5) -1
6)
7)
8) 2| |
9) 1
10)

III) 1) True 2) False 3) True 4) True 5) False


6) False 7) False 8) True 9) True 10) True

: 315 :

You might also like