MEASURE OF

PARTITION/POSITION
• Partition values are positional measures which divides a series into
equal parts say 4 equal parts , 8 equal parts, 10 equal parts or 100
equal parts. The most popular partition values are – Quartiles,
Octiles, Deciles and Percentiles.
• Measures of partition describes the variation or spread in a set of
data.
QUARTILES
Quartiles− The values of a variate that divide the series or the
distribution into 4 equal parts are known as Quartiles. Since
three points are required to divide the data into 4 equal parts
we have three quartiles. 𝑄1,𝑄2 and𝑄3.
• The first quartile(𝑄1), known as a lower quartile, is the value
of a variate below which there are 25% of the observations
and above which there are 75% of the observations.
• The second quartile(𝑄2), known middle quartile, or Median is
the value of a variate which divides the distribution into two
equal parts. It means, there are 50% observation above it and
there are 50% observation below it.
• The third quartile(𝑄3), known as an upper quartile, is the
value of a variate below which there are 75% observations
and above which there are 25% observations. So it is clear
that𝑄1<𝑄2<𝑄3.
Computation of quartiles

• The procedure for computing quartiles is the same as the Median as

explained before.
A) In case of individual and discrete series (After arranging the size of
items in ascending/descending order)
𝑁+1
Q1= =( )th-item
4
2(𝑁+1)
Q2= ( )th-item
4
3(𝑁+1)
Q3= ( )th-item
4
• [II] In case of continuous series (i.e Frequencies with class intervals)
𝑁
−𝑐𝑓
• Q1= 𝐿1 + 4
𝑖,
𝑓
2𝑁
−𝑐𝑓
• Q2= 𝐿1 + 4
𝑖,
𝑓
3𝑁
−𝑐𝑓
• Q3= 𝐿1 + 4
𝑖,
𝑓
Where −
• 𝐿1 a lower limit of a quartile class
• 𝑐.𝑓is a cumulative frequency preceding the quartile class
• 𝑓 is a simple frequency in a quartile class
• 𝑖 Class interval of quartile class
Octiles
• The value of a variate which divides a given series or distribution into 8
equal parts are known as Octiles.
• Each Octile contains 12.5% of the total number of observations. Since
seven points are required to divide the data into 8 equal parts, we have
7 Octiles.
• Computation of octiles
• [I] In case individual and discrete series (after arranging the size of items
in ascending/descending order)
𝑗(𝑁+1)
• 𝑂 𝑗= 𝑡ℎ−item where 𝑗=1 to 7. So if 𝑂4 is required then
8
4(𝑁+1)
𝑂 4= 𝑡ℎ−item.
8
• [II] In case of continuous series (i.e Frequencies with class intervals).
𝑗𝑁
−𝑐𝑓
• 𝑂𝑗== 𝐿1 + 8
𝑓
𝑖,
Where
𝑗= 1 to 7,
𝑐.𝑓− is the cumulative frequency preceding 𝑗𝑡ℎ octile class,
𝑓− simple frequency of the Octile class.
Deciles

• –The value of a variate that divide the series or the distribution into10
equal parts, are called deciles.
• Each part contains 10% of total observations.
• Since nine points are required to divide the data into 10 equal parts, we
have 9 deciles 𝐷1 to 𝐷9. They are called first decile, second decile, e.tc.
The 5𝑡ℎ decile (𝐷5) is the Median.
Computation of deciles

[I] In case of individual and discrete series (After arranging the size of
the item in ascending/descending order).
𝑗(𝑁+1)
𝐷𝑗= ( )𝑡ℎ−item where 𝑗=1 to 9.
10
5(𝑁+1)
For example 𝐷5= ( )𝑡ℎ− item.
10
• [II] In case of continuous series (i.e Frequencies with class intervals) = 𝐷
𝑗𝑁
10
−𝑐𝑓
𝐷𝑗 =𝐿1 + 𝑓
𝑖, where 𝑗=1 to 9, and 𝑐.𝑓−is the cumulative frequency preceding
the 𝑗𝑡ℎ− decile class
Percentiles

• The value of a variate which divides a given series or a distribution

into 100 equal parts are known as percentiles. Each percentile
contains 1% of the total number of observations.
• Since ninety nine points are required to divide the data into 100 equal
parts, we have 99 percentiles 𝑃1 to 𝑃99.
• The percentile 𝑃𝑗 is that value of the variate up to which lie exactly 𝑗%
of the total number of observations.
• For example.
𝑃10= Is the value of the variate up to which lies exactly 10% of
observations, this is the same as 𝐷1.
𝑃25= Is the value of the variate up to which lies exactly 25%
of the total number of observations. This is the same as 𝑄1.
𝑃50= Is value of the variate up to which lies exactly 50% of
total number of observations. This is the same as 𝐷5 or 𝑄2 or
Median. Similarly 𝑃75=𝑄3.
Computation of percentiles
[I] In case of individual and discrete series( after arranging the size of
items in ascending/ descending order).
𝑗(𝑁+1)
𝑃𝑗=( )𝑡ℎ−item where 𝑗=1 to 99.
100
50(𝑁+1)
For example 𝑃50= ( )𝑡ℎ−item where 𝑗=1 to 9.
100
.
[II] In case of continues series(i.e frequencies with class intervals)
𝑗𝑁
−𝑐𝑓
• 𝑃𝑗= 𝐿1 + 100
𝑓
𝑖
Where 𝑗= 1 to 99. 𝑐.𝑓−Is the cumulative frequency preceding the 𝑗𝑡ℎ
percentile class.
example
1.Compute quartiles for the data below.
25,18,30,8,15,5,10,35,40,45
Solution: arrange the data in order. That is
5,8,10,15,18,25,30,35,40,45
• Lower Quartile
=  N  1  11
 th   2.75th
 4  4

Therefore, first Quatile is the 2nd term+0.75(3rd-2th)=8+0.75(10-8)=9.5

examples
2.Given the following frequency distribution table, calculate
Q1,Q3,O5,D8,P70,P10, and P90

marks 0-10 10-20 20-30 30-40 40-50 50-60

No of 10 20 30 50 40 30
student
examples
Find the cumulative frequencies

marks 0-10 10-20 20-30 30-40 40-50 50-60

No of 10 20 30 50 40 30
student
Cumulativ 10 30 60 110 150 180
e
frequecy
MEASURE OF DISPERSION
• A measure of location, such as the mean, median, or mode, only
describes the center of the data. It is valuable from that standpoint, but
it does not tell us anything about the spread of the data.
• In two or more distributions, the central value may be the same but
still there can be wide disparities in the formation of the distribution
• A reason to study dispersion is to compare the spread in two or more
distributions.
Measure of dispersion

• What is dispersion?
Dispersion may be defined as the extent of scattered-ness of items
around a measure of central tendency .
In other words the degree to which the numerical data tend to
spread about an average value is called the variation or dispersion of
data.
• A measure of dispersion indicates the extent to which the individual
observation differ on average from the mean or from any other
measure of central tendency .
• When measure of dispersion is expressed in the units of the variable it
is called “Absolute measure of dispersion”. On the other hand it is
expressed in the form of coefficient ratio or percentage then it is called
“Relative Measure of dispersion.
What is the significance of measuring dispersion
Measures of dispersion are calculated to serve the following purposes.
(1) To determine the reliability of an average - Measuring variability determines the
reliability of an average by pointing out to what extent the average is representative of the
entire data.
(2) To facilitate comparison – Measures of dispersion facilitate comparison of two or
more variability distributions with regard to their variability.
(3) To facilitate control -Measures of dispersion determine the nature and course of
variation in order to control the variation itself.
(4) To facilitate the use of other statistical measures - Measuring variability facilitate the
use of other statistical measures like correlation, regression, statistical inference etc.
Measure of dispersion

• The main measure of dispersion are: (a) Range (b)Quartile deviation or

interquartile range (c) Mean deviation or average deviation (d) Variance
and Standard deviation (e) Lorenz curve.
RANGE
• The range (R) is defined as the difference between the smallest value
and the largest value of series.

• When the data are grouped into a frequency distribution, the range is
equal to the difference between the upper boundary of the highest class
and the lower boundary of the lowest class.
FORMULA
• FOR INDIVIDUAL DATA:
• Range(absolute range)= largest value-smallest value
For continuous data:
Method I:
Range=𝑈𝑙 −𝐿𝑠
Method II
Range=𝑀𝑙 −𝑀𝑠 where
𝑈𝑙 -Upper limit of the largest value class
𝐿𝑠 -Lower limit of the smallest value class
𝑀𝑙 -Mid point of the largest value class
𝑀𝑠 -Mid point of the smallest value class
• For individual observations or discrete series.
𝐿−𝑆
Coefficient of range =
𝐿+𝑆

• For continuous series

𝑈𝐿 −𝐿𝑠
Method I Coefficient of Range=
𝑈𝐿 +𝐿𝑆

𝑀𝐿 −𝑀𝑆
Method II Coefficient of Range=
𝑀𝐿 +𝑀𝑆
Day Monday Tuesday Wednesday Thursday Friday Saturday

Prices (Shs)
2000 2100 2080 1600 2000 2500
• Calculate range and coefficient of range from the following data.

Marks 10 − 20 20 − 30 30 − 40 40 − 50 50 − 60

No. of
students 8 10 12 8 4
(b)Quartile deviation (or Semi-inter-quartile range
This measure of dispersion is based on upper quartile 𝑄3 and the
lower quartile 𝑄1 . The difference between the upper quartile and the
lower quartile is called inter quartile range and one-half of this range is
a quartile deviation.
• Formula
Interquartile range = 𝑄3 − 𝑄1
𝑄3 −𝑄1
Quartile deviation 𝑄. 𝐷 =
2
𝑄3 −𝑄1
Co-efficient of 𝑄. 𝐷 =
𝑄3 +𝑄1
𝑄3 −Upper quartile. 𝑄1 − Lower quartile
• 𝑄. 𝐷 − Is an absolute measure of dispersion whereas
• co-efficient of 𝑄. 𝐷 is a relative measure.
Example.
• Compute the Q.D and coefficient of Q.D from the following data.

Marks 10 20 30 40 50 80

No. of 4 7 15 8 7 2
students
• Compute the Q.D and its coefficient from the following data

𝑥𝑖 10 − 20 20 − 30 30 − 40 40 − 50 50 − 60 60 − 70

𝑓𝑖 12 19 5 10 9 6
Mean deviation

• Mean deviation of a set of observations of a series is the arithmetic

mean of all deviations (without their algebraic signs), taken from its
central value (Mean, Median or Mode).
• In other words it is the average of the modulus of deviation of the
observation in a series taken from the Mean, Median or Mode.
Formula for absolute measure of Mean deviation

• For individual Observations:

𝑁
𝑖=1 𝐷𝑖
M.D =
𝑁
• Where M.D –Mean deviation
• 𝐷𝑖 −Deviations of items from Mean or Median or mode ignoring
the ±signs
• N – Number of observations
• For discrete series( when frequency are given)
𝑛
𝑖=1 𝑓𝑖 𝐷𝑖
M.D =
𝑁
• Where M.D is Mean deviation
• 𝐷𝑖 is deviations of items from Mean or Median ignoring the
±Signs
• N – Number of total frequencies
• 𝑓𝑖 𝐷𝑖 −Product of deviations and the respective frequency
• For continuous series (i.e when both classes and frequencies are given)
𝑛
𝑖=1 𝑓𝑖 𝐷𝑚
M.D=
𝑁
• Where-M.D is mean deviation
• 𝐷𝑚 −Deviations of mid-points of classes from the Mean or Median
• N- Total number of frequencies and 𝑛 −Number of classes.
 𝑀.𝐷 𝑎𝑏𝑜𝑢𝑡 𝑀𝑒𝑎𝑛
• Coefficient of M.D about Mean =
𝑀𝑒𝑎𝑛
𝑀.𝐷𝑎𝑏𝑜𝑢𝑡𝑀𝑒𝑑𝑖𝑎𝑛
• Coefficient of M.D about Median=
𝑀𝑒𝑑𝑖𝑎𝑛
𝑀.𝐷𝑎𝑏𝑜𝑢𝑡𝑀𝑜𝑑𝑒
• Coefficient of M.D about Mode =
𝑀𝑜𝑑𝑒
example
• Calculation of Mean deviation and coefficient of Mean deviation in
case of individual series.
From the following data, calculate Mean deviation and coefficient of
Mean deviation

Roll 1 2 3 4 5 6
No.
Marks 25 55 5 45 15 35
• Calculation of Mean deviation in case of discrete series.
From the following data calculate Mean deviation and coefficient of
Mean deviation.
Marks
5 15 25 35 45 55

No. of Students.
10 20 30 50 40 30
• Calculation of Mean deviation and coefficient of mean deviation in
case of continuous exclusive series
From the following data, calculate the Mean deviation and coefficient
of Mean deviation.
Marks 0 − 10 10 − 20 20 − 30 30 − 40 40 − 50 50 − 60

No. of
students 10 20 30 50 40 30
STANDARD DEVIATION
• Standard deviation is the square root of the arithmetic mean of the
squares of the deviations of all items of the distribution from arithmetic
mean.
• It is usually denoted by the Greek alphabet 𝜎(small sigma).

• It is a measure of how much “spread” or “variability” is present in a sample.

• If all the numbers in the sample are very close to each other, the standard
deviation is close to zero.

• It shows that small standard deviation means high degree of uniformity of

the observations as well as homogeneity of series and vice versa.
• There are two methods of calculating Standard deviation in an
individual series.
a) Deviations taken from Actual mean
b) Deviation taken from Assumed mean
 (deviation taken from actual mean)
• This method is adopted when the mean is a whole number
• [I] Computation of standard deviation in case of individual series when
deviations are taken from the actual mean.
𝑁 𝑑2
𝑥𝑖 −𝑥 2 𝑖=1 𝑖
(Standard deviation 𝜎 = =
𝑁 𝑁
• Coefficient of S.D OR Coefficient of variation is given by
𝜎
• C.V= × 100
𝑋
Variance is given by the square of standard deviation
• From the following data, calculate standard deviation, coefficient of
variation and variance.

Roll 1 2 3 4 5 6
No.
Marks 5 15 25 35 45 55
Computation of standard deviation in case of discrete series when
deviations are taken from actual mean

• Calculate the standard deviation 𝜎 as follows: Standard deviation

𝑛 𝑓 𝑑2
𝑖=1 𝑖 𝑖
𝜎 = 𝑛
𝑖=1 𝑓𝑖
 Calculation of standard deviation, coefficient of variation and variance in
case of continuous series of actual mean.

Marks 0 − 10 10 − 20 20 − 30 30 − 40 40 − 50 50 − 60
𝑋𝑖
No. of
student 10 20 30 50 40 30
s
Deviation taken from Assumed mean
• This method is adopted when the arithmetic mean is fractional value
𝑁 𝑑2
𝑑𝑖 2
• standard deviation 𝜎 = 𝑖=1 𝑖
−
𝑁 𝑁
In case of discrete with frequency or continuous data
Then,
𝑛 2 𝑛 2
𝑖=1 𝑓𝑖 𝑑𝑖 𝑖=1 𝑓𝑖 𝑑𝑖
• Standard Deviation 𝜎 = −
𝑁 𝑁
• Calculate standard deviation, coefficient of variation and variance

Marks 5 15 25 35 45 55
𝑋𝑖
No. of
studen 10 20 30 50 40 30
ts
Marks 0 − 10 10 − 20 20 − 30 30 − 40 40 − 50 50 − 60
𝑋𝑖
No. of
students 10 20 30 50 40 30
THE EMPIRICAL RELATION BETWEEN MEAN, MEDIAN AND
MODE

• A distribution in which Mean, Median and Mode coincide is called

symmetrical distribution.
• A distribution that is not symmetrical, but rather has most of its
values either to the right or to the left of the mode, is said to be
skewed.
• If most of the values of observations in a distribution fall to the right
of the mode then it is said to be skewed to the right or positively
skewed.
• Distributions that are skewed right contain a few unusually large
values of observations. In this case, mode remains under the peak
(i.e., representing highest frequency)
• but the median (value that depends on the number of observations)
and mean (value that is affected by extreme values) move to the
right. The order of magnitude of these measures will be
Mean > Median > Mode.
• But if the distribution is skewed to the left or negatively skewed (i.e.,
values of lower magnitude are concentrated more to the left of the
mode) then mode is again under the peak whereas median and mean
move to the left of mode. The order of magnitude of these measures
will be
Mean < Median < Mode.
• In both cases, the difference between mean and mode is 3 times the
difference between mean and median as suggested by Karl Pearson.
That is,
Mean – Mode = 3 (Mean – Median)
or Mode = 3 Median – 2 Mean

1
Median= [Mode + 2Mean]
3
EXAMPLES
• If the mean and median of a moderately asymmetrical series are 26.8
and 27.9 respectively, what would be its most probable mode?

• In a moderately asymmetrical distribution the values of mode and

mean are 32.1 and 35.4 respectively. Find the median value.