Sei sulla pagina 1di 17

Submitted To

GOVT. COLLEGE OF EDUCATION, JAMMU

Submitted By:- AMRITA BHARTI


Class:- M.ed Ist-Sem
Roll no. :- 24

TOPIC
Biserial coefficient of
correlation

INTRODUCTION
Biserial Cofficient of correlation is designed
for the situation in which we get Continuous
Variable and dichotomous variable.
The term dichotomous means cut into
two parts.
The dichotomy can either be artificial or
natural.
It is in case of artificial dichotomy that
BISERIAL COFFICIENT CORRELATION is
computed.

SOME EXAMPLES OF ARTIFICIAL


DICHOTOMY
The examples of artificial dichotomy are:
Socially adjusted and socially non-adjusted
Successful and unsuccessful
Trained and untrained
Drop outs and stay ins
Conservative and non-conservative etc.

The formula to calculate correlation


by biserial method is-

bi = M
p - Mq
s
T

pq

Where :
Mp

:- Mean of the higher/ superior group

Mq

:- Mean of the lower/inferior


group
:S.D. of the total
T
group/sample.
p & q :- Proportions of the higher or lower group

:- Height of the ordinate of normal curve at point of division


between p & q.

EXAMPLE
Compute the biserial correlation from the
following scores on the performance of
trained group and untrained group in
music competition.
Scores

Trained
group

Untrained
group

Total

70-79

10

60-69

50-59

40-49

16

30-39

13

20-29

11

10-19

10

Solution:
The formula to calculate correlation
by biserial method is-

bi = M
p - Mq
s
T

pq

Where :
Mp

:- Mean of the higher/ superior group

Mq

:- Mean of the lower/inferior


group
:S.D. of the total
T
group/sample.
p & q :- Proportions of the higher or lower group

:- Height of the ordinate of normal curve at point of division


between p & q.

Various steps are involved in the calculation of


biserial coefficient of correlation

STEP-1 :- Identify the higher and lower


group .
i.e. trained group is the higher group and
untrained is the lower group.

STEP-2 :- find the frequencies (f) by


adding frequencies of trained groups
(f) and frequencies of untrained
groups (f) and find the value of N.
scores

Trained group
1
f

Untrained
group f

70-79

10

60-69

07

50-59

08

40-49

16

30-39

13

20-29

11

10-19

10

N 1 = 25

N = 50

N =75

Total( f)

Step-3
Compute the value of mean of the higher/superior
group i.e. trained group:Scores

Trained
group (f1)

x'

fx

70-79

+3

+9

60-69

+2

+6

50-59

+1

+4

40-49

30-39

-1

-4

20-29

-2

-4

10-19

-3

-6

total

N1 = 25

fx = +5

ence, the mean of the trained group Mp = 46.50

Mean of trained grou


Mp = A.M. + fx i
N1
Here, A.M = 44.5 &
i = 10
Now , Mp = 44.5
+5/2510
Mp =
46.50

STEP-4:

Compute the mean value of lower


group i.e. untrained group

Scores

Untrained
group f2

F2x

70-79

+3

+21

60-69

+2

+8

50-59

+3

+12

40-49

30-39

-1

-9

20-29

-2

-18

10-19

-3

-24

Total

N2 = 50

f2x

( A.M. = 44.5 & i =


Mean of untrained group i.e. Mq=A.M. +f2x/N2i
10)
= 44.5 + (-10 = 40.9
5
18)
Thus, the mean of untrained group MQ = 40.9

STEP5 For standard deviation of the total


group ( )
T

Scores

Total

fx

Fx

70-79

10

+3

30

90

60-69

07

+2

14

28

50-59

08

+1

08

08

40-49

16

30-39

13

-1

-13

13

20-29

11

-2

-22

44

10-19

10

-3

-30

90

Total

N= 75

fx = -13

fx =273

S.D. =
fx/N (i = 10)
i
(fx/N)
BY putting
the values
of fx , N , and
(fx) in the above formula then we
get, S.D. = 19

Step6:
Now, proportion of trained and

untrained group (p & q)


p = N /N =25/75 =.33
1

where N1 = total no. in trained group


and
N = total no. in trained and untrained group i.e. 25 + 50 =75
q = N/N =50/75 = .67
Where N = total no. in untrained group and
Hence p= .33, The percent is 33 (trained group)
And
q =.67, The percent is .67 (untrained group)
Assuming the normal curve, the arbitrary cut may be imposed
at the point

STEP-7
Compute u, the height of the normal
distribution curve at the point of
division between trained group or
untrained group.
20/60100= 17%
Area from the mean should be found.
.17 =.362 (from table)
=.362

STEP-8:
Now substituting all the calculated values of Mp, Mq ,
, p, q and u in the Formula i.e.
T

bis

= Mp
Mq

pq

= 46.5 40.9/19 ( .33)(.67)/.362 = .2


(approximately)
Interpretation :
Hence rbis = +.2 shows a very low positive scores on the achievement of
trained and untrained groups in music competition.

CONCLUSION
Biserial correlation can be applied
when we get continuous variable and
dichotomous variable.
Biserial correlation is used in
situation where dichotomy is artificial
in one variable while the other is
continuous
It is an estimate of the product
moment r.

Thank

u
o
y
k
n
a
Th

you

Potrebbero piacerti anche