Professional Documents
Culture Documents
Skewness Practice Materials
Skewness Practice Materials
PART-II
3 MEASURES OF
UNIT SKEWNESS
SYLLABUS
Meaning of Skewnesss - Symmetrical and Skewed Distribution - Measures of
Skewness Absolute and Relative Measures of Skewness Karl Pearson's
Coefticient of Skewness and Bowley's Coefficient of Ske
The measures of dispersion, measures the spread or scatter or variation of the variables
from acentralvalue-Mean. Median or mode. They do not explain the real characteristics of the
distribution. Therefore, to measure the real characteristics of the distribution, Skewness of the
data is calculated.
Bel-Shaped Curve
Asymmetrical Curve
Mo
MEASURES OF SKEWNESS
Skewness
179
lack of
means
asymmetrical, it is called skewedsymmetry (balance
distribution. or equal spread). If a distribution is
distribution forms a bell-shaped curve when the values are equal. The symmetrical
When the value of Mean is more
plotted on a graph sheet.
than the
Positive Skewness. When the value of Mode is valuethan of Mode (When X>Z), it is called
Negative Skewness. greater Mean (when X>ZI, it is called
Definitions
Garret defined "A distribution is said to be 'skewed' when the
mean and the median fall at
different points in the distribution, and the balanace center of
(or gravity) is shifted to one side or
the other-to left or right."
According to Croxton and Cowden "When a series is not symmetrical, it is said to be
asymmetrical or skewed".
Morris Hamburg defined, "Skewness refers to the
shape of a frequency distribution."
or lack of asymmetry symmetry in the
Tests of Skewness
To ascertain the presence
of skewness in a distribution, the following tests may be applied.
Skewness is said to be present if:
(i) The values of mean, median and mode are different;
(i) When plotted on a graph, the frequency distribution do form
Curve;
not a bell-shaped
(ti) The sum of the positive deviations from the median is not equal to the sum of the
negative deviations from the median;
(iv) Quartles Q and Q3, are not equi-distant from the median.
Skewness is absent when the above conditions are not satisfied.
Difference between Dispersion and Skewness:
(1) Dispersion or variation is concerned with the amount of variation rather than with
the direction. Skewness tells about the direction of the variation.
Measures of
skewness depends upon the amount of dispersion.
(2) Dispersion is widely used statistical tool in decision-making, whereas, skewness is
used rarely.
Measures of Skewness
(i) Absolute Measures of Skewness : Skewness can be measured in absolute terms
taking the difference between Mean and Mode. That means,
by
Absolute Skewness = X-Mode
When skewness is based on quartiles, absolute measure of skewness is calculated
by usingthe formula:
Absolute Skewness Q3+Q1-2 Median.
Ifthe answer is positive, it means that there exists positive skewness (X>Z) and if
the answer is negative, there exists negative skewness X < Z).
cOEFFICIENT OF SKEWNESS
KARL PEARSON'S
or simply Pearsonian
coefficient of skewness (Sk.)
Karl Pearson's coefficient of skewness
mean and mode. This
difference is divided by Standard
is based on the difference between
Deviation to relative measre. That is :
get or 3m-Mu
Skp
Mean- Mode X-2
Standard
empiricalDeila, Mean]
MLA 3Mat
In the absence of mode, the [Modesoskewness is
formula, [Mode = 3 Median -2
calculated
3M-2f
using the
is used instead. In such a case, Pearson's coefficient of
formula:
3 m u
skp X-z
Calculation of Mean and Standard Deviation
(X-X)
40 -10 100
36 -14 196
42 -8 64
53
45 25
50 0
68 18 324
61 11 121
55 5 25
EX=450 2x864
MEASURES OF SKEWNESS
181
450
= 50
9
864
N 9 = 9.798
5th item
Median 50
Sk
3(X Me)
Sk, 3(50-50) = 0
9.798
Thus, there is no skewness in the distribution.
Example 2. The wages paid to 550 workers per day in a manufacturing unit is as follows.
Wages (Rs.) 100 200 300 400 500 600 700 800 900
No. of Workers 35 40 48 100125 87 43 22 50
Calculate Karl Pearson's coefficient of skewness.
Solution.
Sk, =
Calculation of Mean and Standard Deviation
(x-500
100
100 35 4 140 560
200 40 9 120 360
300 48 96 192
400 100 100 100
500 125 10 0 0
600 87 87 87
700 86 172
800 22 9 66 198
900 50 16 200 800
N=550 fd' =- 17| Efd2=2469
=A+ N xi
500+ x 100=496.91
-3
N
xi
2469 x100
550 550)
= V4.489-0.00095 x 100 =
4.48805 x 100
2.1185 x 100=211.85
Since the maximum frequency is 125, Mode
=
550- 500
Skp - z
496.91-500-0.01458
21 185
Example 3. Calculate Karl Pearson's Coefficient of skewmess.
No. of Children 0 1 2 3 5 6
No. of Families 10 16 25 18 11 8
Solution. Skp
Calculation of Mean and Standard Deviation
d
0 3 9 21 63
1 10 -2 4 20 40
2 16 16 16
25 0
18 18 18
11 2 4 22 44
6 8 3 24 72
N 95 2fd=7 2fa=253
X =A+ N
=3+ 95 =3.0737
fd
N N
=
2.663-000543 =
/2.65757
= 1.6302
MEASURES OF SKEWNESS
183
Mode =3.
Sk,X-Z 3.0737-3-0.04519
L.6302
Example 4. Calculate mean wage and Karl Pearson's
following: coefficient of skewness from the
25 men get the rate of
32 men get the rate of 50 per hour
40 men get the rate of 55 per hour
75 men get the rate of 60 per hour
102 men get the rate of 65 per hour
70 per hour
60 men get the rate of
26 men
75 per hour
get the rate of 80 per hour
Solution. Take wages per hour as X and
number of men as frequency
Calculation of Mean and Standard Deviation
X
fd
(X-70)
50 25 - 20
400 500 10000
55 32 - 15
225 480 7200
60 40 - 10
100 -400 4000
65 75 25 -375 1875
70 102 0 0 0
75 60 5 25 300 1500
80 26 10 100 260 2600
N=360
fd-1195 f-27175
A+ 70+
N =66.68
360
-2fdEfd 27175(-1195
N 360 360
=75.4861-11.0187 = 64.4674 =8.0292
The highest frequency is 102,
..
Mode = 70
Sk, Z6.68-70
8.0292
-0.4134
Example 5. Calculate Karl Pearson's coefficient of skewness from the
following table.
| Lifetime (in hours) 300-400 400-500 500-600
No. of bulbs
600-700 |700-800 800-900 900-1000
25
56 60
75 48
30 15
Solution.
300-4000 350
100
3
400-500 450
9 25 -75
2 225
500-600 550
4 56 - 112
1 224
600-700 fo> 60 - 60
650 0 60
700-800 750
A775 0
0
800-900 850
f 48 48
2 48
900-1000 950
30 60
3 9 120
15 45
135
N-309 2fd'=-94 fd2=812|
=
2.6278-0.09254 x 100
=
25353x 100 1.5923 x 100
= 159.23
Since, highest frequency is 75; modal class is 600-700
Mode L+|2f-fo- $2
600+ 75-60
x 100
2(75)- 60-48
600+x100 600+35.714
Mode = 635.714
Sk
Skp -Z 61958-635.714
159.23
Sk =-0.1013
Example 6. Find Pearson's coefficient of
skewness from the
following data:
Wages (Rs.) 270-280 280-290 290-300 300-310 310-320 320-330 330-340 340-350
No. of Workers 12 18 35 42 50 45 20 8
Solution.
Skp -2
270-280
(mid-polnt) -305)
d'2
I00 d fd'2
275
280-290 285
-3
290-300 295
-2 12 - 36
4
300-310 1 18 108
305 1 36
310-320 0 35 72
315 0 35
So 42 35
320-330 325 0
330-340 50 50
0
335 Sa45 50
340-350 3
9 90
345 20 180
16 60
180
32
128
N-230 2fd'=125
X A+xi305+610 | Zfd2=753
310.4348
NN 2fd(fdiXi 23020X10
= x10 230 230 X10
=
3.2739-029537
= 1.72584x 10
x 10 =
J297853 x 10
o =17.2584
Since, highest
frequency is 50; mode lies in 310-320
class.
Mode L
310+6.154
Mode 316.154
SkX-2 310.4348-316.154
172584
Sk =-0.3314
Example 7. Find Pearson's coeficient of
skewness from the
90-100 100-110 110-120 following data:
|120-130 130-140 140-150 150-160
f 4
2221 19 10 3|160-170170-180
2 18
Solution. 1
skp
J0-100 T0
95
100-110 105
-4 16
-16
110-120 115
-3 9 2
64
-6 4
120-130 125
-2 4 18
18
6
-36
130-140 135 1 22
72
24
-22
140-150 145
0 0 21 0
22
46
150-160 155
1 1 19 19
0
67
160-170 165
2 4 10
19
86
20
170-180 175
3 9 3 9
40
96
4 16 27
1 4 99
16
N=100 2fd'=-28fd'2=278| 100
X
=A+ xi=135 + 100 x1o=135-2.8 =132.2
fa2 (fa' xi = 278 282
N N x 10
100 100
o
2.780.0784 x 10 /2.7016 x10
= 16.436
= =
1.6436 x 10
As,
table andfrequencies are concentrating around 110- 150, it is
analysis table. necessary to prepare grouping
Grouping Table
X Col.
Frequencies
Col.I Col.II Col.IV
90-100 4 Col. V Col.V
100-110 2
24
110-120 20
18
42
120-130
40
22
130-140 43 61)
21
62
40
140-1500 19
50
29
150-160 10
32
13
160-170 3
14
170-180 1
3
Median 2 Mean. bimodal in nature.
Median =
size of Nth
2 item= size of 100th
2
item
= 50th item
Therefore, Median lies in 130 -
140 class
N
Exact Median =L+ C.f.
x i
50-46
= 130 +1x10130+
21
Median =131.904
Sk 3X-Med) 3(1322-131904)
16.436
Skp 0.054
Example 8. Calculate Karl Pearson's
coefficient of skewness from the
Weekly Wages (Rs.) No. of Workers
following data.
100 and Above
50
200 and Above
46
300 and Above
36
400 and Above
18
500 and Above
6
600 and Above
3
700 and Above
Solution. Convert open-end classes into
normal classes and then solve
skp X-2
=A xi-450+x100
370
a = 128.06
f-foxi
ModeL* 2f-fo-f2
18-10
300+ -x 100
2 2 (18)-10-12
300+x100=300+57.143
Mode =357.143
Sk X-Z 370-357-143
skp 128.06
Skp =0.10039
Example 9. Compute Karl Pearson's coefficient of skewness from the follwoing data:
Variable21-23 24-26 27-29 30-32 33-35 36-38
Frequency 17 193 399 194 27 10
Solution. Here, variable are given in inclusive form. Convert them into
exclusive form
and then compute Mean, Mode and Standard Deviation.
Skp X-Z
(mld-point)
(m-28)
3
20.5-23.5 22 -2 4 17 -34 68
23.5-26.5 25 -1 So 193 -193 193
26.5-29.5 28 0 A399 0
29.5-32.5 31 1 S2 194 194 194
32.5-35.5 34 2 4 27 54 108
35.5-38.5 37 3 10 30 90
N=840 fd=51fd-653
=A -28 840
28.18
2 (fdxi 653 (51
NN) =40 840
x3
Sk 2 2818-28
2.6388
Skp =0.0682
Example 10. Consider the following data:
Distribution A Distribution B
Mean 100 90
Median 90 80
Standard Deviation 10 10
State whether the following statements are true or
false.
() Distribution A has the same degree of the variation as
Distribution B.
(i) Both distributions have the same
degree skewness.
of
Solution. ( Coefficient of
variation =
x100
10
x 100 ==10%
Distribution A: 100
10
Distribution B: 100 11.11%
X
O0
=
o 4.472
Solution. skp
3X Me)
a = Variance = V10 =10.4881
3(59.5-55.7)
Skp 10.488
191
MEASURES OF SKEWNESS
114
Sk
104881
= I.0869
Example 14. Given C.V. =30%: Mean 25: Mode = 16. Find Sk.
Solution. CV X100
30 =x 100
5
25x 30
7.3
100
3(X Me) 3(25- 161
Skp = l.2
Example 15. In a moderately skewed frequency distribution the mean is 20 and the
Coefficient of skewness
median is l8.5. If the coefficient of variation is 30%, ind the Pearsonian
of the distribution.
30 1
a 6
320-185
Skp 0.
Example 16.The following information was obtained from the records of a factory relating
to wages