Measurement of Central Tendencies
1 Central Tendency
An average value generally lies in
central part of the distribution and
such values are called the measurement
of central tendency
Following are the measures of central tendang
a Arithmetic mean
2 Geometric Mean
31 Harmonic mean
a weighted Arithmetic mean
Combined Mean
6 Median
7 Mode
181 Relation btw A M G M Hm
Quartile decile
Central
tendency
It can be divided into 2
groups
a mathematical Average
a Arithmetic mean
b Geometri mean
a Harmonia mean
121 Position Average
9 Median
b mode or Position Average
2 Arithmetic mean
Individual Observation
be observations
If Xi x2 x2 xn n
then AM is given by
I At X2 t Xn
n
ie I Ex
if
Ex Find A M of 174 175 179 172 173
I 174 1754179 172 173
1711
5
Eg The mean of 30 terms is 8 One of the
term amongst the series is 12 which
we found Then find the correct
mean if
wrong
a The
the
wrong number is subtracted from
series
b Wrong number is replaced by 18
Sof I 8 7 30
Exe 8 30 240
I
II
240 12
Forrey 29
b
Icozzect 240 12
30
18
O find the A M of first n natural nos
I 1 2 th
5 htt
Find the Am natural
a of squares of n
numbers
I 12 A Gtl
24
2nd
G
I Discrete frequency Distribution
I fix taxa t fran
fi tta t tfn
of
fix
I
y
N Eti
0.1 Find the A m of distributed frequency
distribution
Ki 6 7 10 13 16 19
fi 7 10 15 20 35 30
ki fi fix
4 7 28
7
10
10
IT
70
I TO
I
fi
13 20 260
16 35 560 L
19 30 570
Efi 117 Efik 1638
1 Shortcut Method Assumed Mean method
If the values of se and f are large
then take the deviation from arbitrary
point A
I At Efidi
Efi
where A Assumed Mean
di sci A deviation from each
term
ki fi di ki A fidi
y 7 g 63
7 10 G 60
5 134
If
10 15 3 45 x
A 713 20 0 6
16 35 3 105
19 30 6 180
Eti 117 Efidi 117
Step Deviation Method
Mi i 1,2 n
KEA
mean I At h E fini
Eti
3
Weighted Arithmetic mean
iii l
Weighted Arithmetic Mean
Wix Wax t t
Waxy
Wit Wa t f Wn
I Witi
Wi
0.1 Weighted A m of first n natural nos
when weight of each natural number is
proportional to itself
sod
I 1 It 2.2 3.34 n n
1 2 th
hints
6 Cht
lentil
n
2
231
47 Combined Mean
If Ta ta I the mean of K
are
series of sizes hi na nk respectively
then the mean of composite series is
given by
I N T t hat t next
N that the
9.1 Frequencies of the rallies 0,1
2,3 D
are Nco Ny Ny Men Then its mean
is
I 0 no 1 No 2 Ng t th No
motha t
then
rna
I E
27
E H
ng ca
2n
n
tooth gt
In on
2
Is Iz
0.2 Mean of 12 22,33 n is 467 11 Then
find n
Rt 22h th 4
n
none anti
Gn
497
1 27 32 1 276 N
Nell
0 If a variable x is expressed as a linea
function of 2 variate V and V in the form
of X gut bu then meant of x be
X a Ut br
9
nd In by
I a T t bit
5
Properties of Arithmetic mean
In a statistical data the sum of
the deviation of items from a m is
always cero
i e Gi I 0
OI
EE Xi ni o
2 In a statistical data the sum of
squares the
of deviation of items from
Am is least i e
Gi IT is least
3
If each of the n given observation
be doubled their mean i's doubled
4 If I is the mean of Xi Xa xn
The mean of ax saxa axn is at
where a is any digit different from
zero
Arithmetic mean is independent of
i It is not
origin e
effected by
any change in the origin
G Meanof observation of group is t If
every observation is divided by a and
increased by 10 Then Mean of new
observation is
Sof Kit Kat tan I
n
Egitto 11 10 t
Hatto
n
Kitkat ten
lift ta n
II 10
O Mean of K K2 23 In is I
observation xp is replaced
If
by Xp Then
new mean is
KitKat t Xp t ten I
n
Kitkat t an ME
xp
Icorrect MI Sep tap
n
Geometric Mean om
d Gm of Individual Data
If Ki ka s values of
kn ar n
a variate x none of them being zero
then Geometric Mean u is defined as
be X h
sea Sen
12 G m of Grouped data
let Ki ka Ks en be n observations and
let fifa th be their corresponding frequent
then their G in is
YN
G Kt cat oath
where N
Eti
Eg Find U M
2 4 8 16
z 3 2 I N 10
b m
2443.83161410
22 4
4 Harmonic mean
It is reciprocal of mean of
reciprocal
A Individual observation
H n
I It ten
n
i e
H
IE i
2 Him of grouped data
let Xi x2 Xn be n observation and
let fr fa th be their corresponding
frequency then Him is
H Efi N
E Cf tf t fan
Relation btw Am Gm and H M
A M I G m I H M
Equality sign holds when all the
only
observations in the series are same
5
Median
It is the
middle must or the
central value of the variate in a set of
observation when the observations are
arranged in either ascending or
descending order of their magnitudes
It divides the arranged series in
two parts
4 Median of Individual series
a
If n is odd then
th
in Value of observation
g
b n is even
If then
in Mean of
f hand Iti observation
ie M E observation 2 tithobservation
2
2 Median of discrete frequency distribution
Algorithm to find the median
step g find the cumulative frequency CF
Step 2 Find
I where N
Effi
step 3 See the c F just greater than Nz
and determine the corresponding
value of the variable
step 4 The value available in step 3 is the
value of the median
Ex Find median for distributed frequency
distribution
Ki fi C F
I 8 8 N 120
2 10 18
Go
3
4
11
16
29
45
I
1 5 20 65 1 Median 5
6 25 90
7 15 105
8 9 11 4
g 6 120
Efi 110
3 Median of grouped data or continuous
series
let the no of observation be n
4 Prepare the cumulative frequency
table
Find the median class i e class in which
th observation lies
an The median value is given by
formulae
Median M It 1 F th
f
N Total frequency Eti
e lower limit of median class
f frequency of the median class
F cummulative frequency of the
class preceding the median class
h class interval width of the median
class
Ex Find median of grouped 1 continuous
data distribution
Range fi c f Efi N 24
0 10 5 5 12
10 20 4 g I
l 20 30 8 17 1 20 h 10 F g
30 40 7 24 f 8
M 20 1289 to
20 t
3 10 23 75
6 Mode value in a series
Mode is that
which occurs most frequently
In a frequency distribution mode is
that variate which has the maximum
frequency
x fi
eg
2
3 2
65
6 Mode
117 Computation of Mode
Mode for individual series the case In
of individual series the value which
is repeated maximum number of times is the
mode of the series
His Mode for grouped data discrete
frequency distribution series
In the case of discrete frequency
distribution first arrange value of variate
in ascending or descending order and their
respective frequency against them Now find
the maximum frequency value
Mode for continuous frequency distribution
is First find the the class
model class i e
which maximum frequency The model
has
class can be determined either by inspecting
or with the help of grouping data
The mode the formula
Ci is given by
Mode lt fm fm h
Ifm fm fm
where
l lower limit of the Model class
h width of the model class
fms frequency ofof themodel class
fm frequency class preceding
Model class
fmy frequency of the class succeeding
model class
G1 Find mode of grouped 1 Continuous data
distribution
Range fi
es
10 20
Imitm
4
30 40 F fmtl
4 10
It fm fm h
Mode
2fm fm fm
20 8 4 X 10
16 y y
22
7 Measures of Dispersion Dispersion is the
statistical distribution
in measures of deviation of its value about
the value
average
The measures of dispersion commonly used
are
in Range
tis Quartile deviation or the semi
inter quartile range
did Mean Deviation
in Standard Deviation
Mean Deviation
Mean deviation is defined as the arithmetic
mean of the absolute deviations of all
the values taken about any central value
Mean deviation of individual observations
If Xi x2 xn are n values of a
variable x then the mean deviation
from an average A median or AM is
given by
M D I É Ki Al
In El dit when di Xi A
lil Mean deviation of a discrete frequency
distribution
x
If x2 Xn are n
observation with frequencies fi ta th
then mean deviation from an average
A is given by
Mean deviation
Eti ki Al
where N
Effi
Eg Find mean deviation from median of
47 50 58 45 53 59 47 60 49
SI 45 47 47 49 5111 53 58,59 60
MD 5 3 3 1 T 0 3 849 10
g
8 Variance and standard deviation
The variance of a variate x is the
arithmetic mean of the square of all
deviations of x from the arithmetic
mean of the observations and is denoted
2
Varese
by The or
root of the
positive
of
square
is
variance a van'ate x Kla standar
deviation i e standard deviation
Fare
111 Variance of Individual observations
If Xi x2 Xn are n values of a
variable x then by definition
varix
I E xi IT a 0
2
or Varix
I Xi I
If values of a variable x are large
the
the calculation of variance from the
above formula is quite tedious and time
consuming In that case we take deviation
from an arbitrary point A says then
Varix
In Indi t E di iii
in Variance of a discrete frequency
distribution
If x x2 x are n observation
with frequencies fi ta tn then
Varix
E ti Lxi Is is
Varix tix I Ii
I
If the values of x or f are
large
we take the deviations of the values
of variable x from an arbitrary point
A say
di Xi A i 1,2 n
Vara
E fid Éfidi
where N
Eti
standard deviation
SD Varix
Eg Find Variance of first n natural nos
varix
Eff E
Ki 1,2 3 n
nugent Ed
Inti Ntl
Ynet 4
anti
of ng
4 1 47
2,237
3
htt n 1
2,31
12