Documente Academic
Documente Profesional
Documente Cultură
EDUCATION
JAYDEV BIHAR, BBSR
BUSINESS MATHEMATICS AND STATISTICS
STANDARD DEVIATION
Teacher’s Signature:
Name of subject teacher: Judhistir das
ACKNOWLEDGEMENT
I would like to express my gratitude to my
teacher Mr. Judhistir das who gave me the
amazing opportunity to do this wonderful
project report on Standard deviation.
Secondly, I would also like to thank my
parents and friends who helped me a lot in
finalizing this project within the stipulated
time frame.
1
Merits and Demerits of Standard Deviation:
Merits:-
1) It is rigidly defined and free from any
ambiguity.
2) Its calculation is based on all the observations
of a series and it cannot be correctly calculated
ignoring any item of a series.
3) It strictly follows the algebraic principles and it
never ignores the + and - signs like the mean
deviation.
4) It is capable of further algebraic treatment as it
has a lot of algebraic properties.
5) It is used as a formidable instrument in making
higher statistical analysis viz. : correlation
skewness regression and sample studies etc.
6) It is not much affected by the fluctuations in
sampling for which it is widely used in testing the
hypotheses and for conducting the different tests
of significance viz. : x test t-test etc.
2
7) In a normal distribution x̅ +1 6 covers 68.27%
of the values for which it is called a standard
measure of dispersion.
8) It exhibits the scatter of dispersion of the
various items of a series from its arithmetic mean
and thereby justifies its name as a measure of
dispersion.
9) It enables us to make a comparative study of
the two or more series and to tell upon the
inconsistency or stability through calculation of
the important factors viz. co-efficient of variation
variance etc.
10) It enables us to determine the reliability of the
Mean of the two or more series when they show
the identical means.
11) It can be calculated through a good number of
methods yielding the same results.
4
Demerits:-
1) It is not understood by a common man.
2) Its calculation is difficult as it involves many
mathematical models and processes.
3) It is affected very much by the extreme values
of a series in as much as the squares of deviations
of big items become proportionately bigger than
the squares of the smaller items.
4) It cannot be used for comparing the dispersion
of two or more series given in different units.
5
Methods Of Computation:
Basing up the definition and the essential
characteristics cited above there can be four
different methods for computing the standard
deviation of a series.
These are as follows:
1. Direct method
2. Short cut method
3. Step deviation method
4. Method based on values
6
1. Direct method -Under this method the
standard deviation is computed on the basis of
the deviations of the items from their actual
arithmetic average. This method is suitable when
the arithmetic average of the series comes out to
be a round and definite figure.
The formulas applied under this method are:
In case of individual In case of discrete and
series Continuous
(σ =√∑x2/N) series (σ =√∑fx2/N)
7
Example of Direct method (Individual Series)
X: 5 10 15 20 25 30 35
X (X- x̅) x2
x
5 -15 225
10 -10 100
15 -5 25
20 0 0
25 5 25
30 10 100
35 15 225
Total 140 _ 700 N=7
x̅= ΣX/N=140/7=20
σ =√∑x2/N = σ =√700/7= 10
Hence, the standard deviation of the given
observation is 10.
8
Example of Direct method (Discrete Series)
X 3 5 7 10 12 15 20 22 25
F 5 10 4 8 5 3 9 6 10
9
Example of Direct method (Continuous Series)
Class 0-10 10- 20- 30- 40- 50-
20 30 40 50 60
Frequency 11 8 4 9 8 10
10
2. Short-cut method -Under this method, the
standard deviation is computed on the basis of
the deviations of the items takes from assumed
arithmetic average. This method is suitable where
the actual arithmetic average of a series is found
to be a fractional or an infinite figure.
The formulae used under this method are as
under:
In case of Individual In case Discrete and
Series Continuous Series
σ = √[(∑d²/N) – (∑d/N)²] σ = √[(∑fd²/N) –
(∑fd²/N)²]
11
∑fd2 = sum of the products of the squares of
deviations from the assumed mean and their
corresponding frequencies.
∑fd= sum of the products of the deviations from
the assumed mean and their corresponding
frequencies.
∑f = total of the frequencies or N
12
Example of short-cut method (Discrete series)
X F (X-A) d2 fd2 fd
D
8 1 -7 49 49 -7
9 2 -6 36 72 -12
15 3 0 0 0 0
23 5 8 64 320 40
5 6 -10 100 600 -60
11 7 -4 16 112 -28
19 8 4 16 128 32
8 9 -7 49 441 -63
10 10 -5 25 250 -50
12 11 -3 9 99 -33
Total N=62 2071 -181
σ = √ [(∑fd²/N) – (∑fd²/N) ²]
σ = √ [(2071/62) – (-181/62)²]
σ = √ [3.40-(-2.92) ²]
σ = √ 33.40-8.53
σ = √ 24.87= 4.99 (approx)
13
Example of short-cut method (Continuous series)
Weight: 115- 125- 135- 145- 155- 165-
125 135 145 155 165 175
Frequency: 4 5 6 3 1 1
σ = √[(∑fd²/N) – (∑fd²/N)²]
σ = √(3700/20)-(-50/20)2
σ = √185-6.25 = √178.75= 13.37(approx)
Hence, the standard deviation of distribution is
13.37
14
3. Step deviation method -This is an extension of
the shortcut method explained above. Under
method the standard deviation is computed on
the basis of the assumed deviations after they are
reduced by the greatest possible common factor.
However the derived result is again multiplied by
the said common factor to arrive at the desired
value of the standard deviation. This method is
suitable only when the assumed deviation figures
appear to be of big size and all of them can be
divided by a common factor.
The formulae under this method are thus
modified as under:
In case of Individual In case Discrete and
Series Continuous Series
σ = √[(∑d’²/N) – (∑d’/N)²]x σ = √[(∑fd’²/N) –
C (∑fd’²/N)²]x C
15
Example of step deviation method (Individual series)
16
Example of step deviation method (Discrete
series)
X F (X-A) d/c d’² fd’ fd’²
d d’
14.5 3 -30 -3 9 9 27
24.5 61 -20 -2 4 -122 244
34.5 223 -10 -1 1 -223 223
44.5 137 0 0 0 0 0
54.5 53 10 1 1 53 53
64.5 19 20 2 4 38 76
74.5 4 30 3 9 12 36
Total N=500 -2510 -251 659
17
4. Method based on values- Under this method
the standard deviation is calculated on the basis
of the values of the exact items as the assumed
mean is here considered as zero and the
deviations are ultimately the exact values. The
formula used here are quite analogous to that of
short-cut-method. This method is suitable and
advisable only where the size of the items is short
and simple.
The formulae used under this method are as
follows:
In case of Individual In case Discrete and
Series Continuous Series
σ = √[(∑X²/N) – (∑X/N)²] σ = √[(∑FX²/N) – (∑FX/N)²]
18
Example Method based on values (Discrete
series)
Vales F X FX FX²
0-2 2 1 2 2
2-4 7 3 21 63
4-6 12 5 60 300
6-8 19 7 133 931
8-10 9 9 81 729
10-12 1 11 11 121
Total 50 308 2146
19
Example Method based on values (Continues series)
Value 0-2 2-4 4-6 6-8 8-10 10-12
Frequency 2 7 12 19 9 1
Vales F X FX FX2
0-2 2 1 2 2
2-4 7 3 21 63
4-6 12 5 60 300
6-8 19 7 133 931
8-10 9 9 81 729
10-12 1 11 11 121
Total 50 308 2146
20
Relative Standard Deviation
Like any other measure of dispersion, standard
deviation can also be measured in two ways viz.:
(1) Absolute
(2) Relative
Absolute standard deviations are computed in the line
discussed above. But they are not suitable for making a
comparative study of more than one series. For this
purpose relative standard deviations are to be
computed. A relative standard deviation is again of two
types.
They are:
1) Co-efficient of standard deviation. This is expressed
as the ratio of the absolute standard deviation to the
arithmetic average of the series. This is represented as
under:
Coefficient of S.D= Standard deviation/Mean or σ= σ/
x̅
21
2) Co-efficient of variation. The co-efficient of standard
deviation thus expressed does not give the answer in a
convincing manner. To get the answer in a more
convincing manner, co-efficient of variation is found
out which is expressed as the percentage of standard
deviation of the arithmetic average of the series. Thus
it is represented as under:
C.V. = (σ/x̅) x 100
The co-efficient of variation guides us in ascertaining
the relative degree of variability or stability of a series.
A series with more co-efficient of variation is regarded
as less stable or less consistent than a series with less
co-efficient of variation. The C.V. is a pure number and
is independent of the units of measurement and is thus
suitable for comparing variability, homogeneity or
uniformity of distributions.
22
Algebraic Properties of standard deviation
Standard deviation has a lot of algebraic properties for
which it is capable of further algebraic treatment and
considered as the best of the measures of dispersion.
Some of its algebraic properties a depicted here as
under.
23
4. In a symmetric and moderately asymmetric series it
maintains the empirical relation with the other
measures of dispersion as under:
a Q.D =(2 σ /3)
b M.D =(4 σ /5)
c Range =6 σ
d Q.D =(5/6)M.D
24
Thus, the standard deviation of the first 5 natural
numbers would be
σ5t=√1/12(5²-1= √1/12(24) = σ=√2=10414
7. It is possible to determine the Combined Standard
Deviation of any two or more groups
25