STATISTICS I

CHAPTER 3: MEASURE OF DISPERSIONS

DHIRAJ GIRI
KATHMANDU UNIVERSITY
SCHOOL OF ARTS
2025
Are rainfall in three stations A, B, and C alike?
Annual Rainfall in Different Stations Mean Median Mode
Station A: 50 50 50 50 50 50 50 50 50
Station B: 48 49 50 50 51 52 50 50 50
Station C: 42 46 50 50 54 58 50 50 50

• A, B and C have same central value but scatteredness of

items differs widely from one another.
• Series A is stationary no variation, Series B is less
dispersed and Series C is more dispersed from central
value.
• Series B is more homogenous than series C or series C is
more heterogeneous than series B.
• Only central value is not sufficient to make comparison of
series.
Measures of Variation
• Mean and Median only describe the central location, but not
the spread of the data.
• Two distributions may have the same central value, but
different variability
• Statistics that describe variability are called measures of
spread/variation/dispersion
• Variation is the amount of spread or scatteredness
in the data.
• The scattering of the values of a variable from
their average (central value) is known as
dispersion.
• Dispersion measures the extent to which items vary from
central value.
• Homogeneity refers to the degree of similarity within a set of
data. The more homogeneous a set of data is, the better the
Why to Study Measure of Variation
• To increase our understanding of the pattern of the data.
• To have an idea about the variability (i.e. homogeneity or
heterogeneity) of the given distribution.
• To judge the reliability of our measure of central tendency.
If the data are widely dispersed, the average or central location is less
representative.
• Measures of dispersion enable a comparison to be made
of two or more series with regard to their variability.
A measure of dispersion may be expressed in an Absolute
form or in a Relative form.
 Absolute Measure:
• Are expressed in terms of the original unit of a
series.
• It states the actual amount by which the value of an
item on an average deviates from a central value.
 Relative Measure: (Coefficient of Dispersion)
• It is the ratio of absolute measures of dispersion to
suitable average, and are thus a pure number,
independent of unit of measurement.
• Are used for comparing the variability of two or
more than two series, even if the series are of
different units of measurement.
Types of Measures of Dispersion:
Measures of variation give information on the spread or
variability or dispersion of the data values.
(i) Range
(ii) Inter-quartile Range
(iii) Quartile Deviation
(iv) Mean Deviation and
(v) Standard Deviation.
Measures of Variation: The Range
• Simplest measure of variation
• Difference between the largest and the smallest values in
the series:
Range  X largest  X smallest
Example:

• The range measures the total spread in the set of data.

Why the Range Can Be Misleading?
• Does not account for how the data are distributed between
the smallest and largest values.

• Sensitive to outliers
1,1,1,1,1,1,1,1,1,1,1, 2, 2, 2, 2, 2, 2, 2, 2,3,3,3,3, 4, 5
Range = 5 − 1 = 4

1,1,1,1,1,1,1,1,1,1,1, 2, 2, 2, 2, 2, 2, 2, 2,3,3,3,3, 4, 120

Range = 120 − 1 = 119
L-S LS
Coefficient of Range  Mid Range 
LS 2
Partition Values
The partition values are those which divide the whole
observations into number of equal parts. There are three
different partition values.
I) Quartiles
II) Deciles
III) Percentiles
Quartiles
• Quartiles are the three points which divide the distribution
into four equal parts.
• There are three quartiles Q1, Q2, Q3 respectively called
lower, middle (median) and upper quartiles.
25 % 50% 75% 100%
0%

L o w est Hig h e st
O b s e r v at i o n Q2 Q3 O b s e rv a ti o n
Q1

Md

th
 n 1 
Q1 Lower quartile (First quartile) Value of   item
 4 
th
 n 1 
Q 2 Middle quartile (Second quartile) Value of 2  item
 4 
th
 n 1 
Q 3 Upper quartile (Third quartile) Value of 3  item
 4 
Deciles
• Deciles are the nine points which divide the distribution
into ten equal parts.
• Namely there are nine deciles D1, D2, D3,…, D8, D9
0 % 1 0 % 2 0 % 3 0% 4 0 % 5 0% 6 0 % 70 % 8 0 % 9 0% 10 0 %

Lo w est Hig he st
O bse rva tion Ob serva tio n

Md th
 n 1 
D1 First decile Value of   item
 10 
th
 n 1 
D2 Second decile Value of 2  item
 10 
th
 n 1 
D 9  Nineth decile Value of 9  item
 10 
Percentiles
• Percentiles are the 99 points, which divide the distribution
into hundred equal parts. Namely there are ninety-nine
percentiles,1 %P21%, P2,…………. P98, P99
0% 50 % 9 8% 99 % 1 00 %

Lo w es t H ig h e s t
O b s e rv a tio n O b s e r va ti o n

Md
th
 n 1 
P1 First percentile Value of   item
 100 
th
 n 1 
P2 Second percentile Value of 2  item
 100 
th
 n 1 
P99  Ninetynineth percentile Value of 99  item
 100 
How to compute quartile, deciles and percentiles in case
of grouped frequency distribution
N
i    cf
 4
Qi  L  h for i 1, 2, 3
f
N
i    cf
 10 
Di  L  h for i 1, 2 , ...,9
f
 N 
i    cf
 100 
Pi  L  h for i 1, 2 , ...,99
f
Quartile Deviation or Semi-Interquartile Range
• A measure of dispersion based on quartiles is known as
Quartile Deviation or Semi-Interquartile Range.
• Inter-Quartile Range = Q3 - Q1
• Half of the inter-quartile range is called Semi-Interquartile
Range, which is also known as Quartile Deviation (QD).
Q3  Q1
QD 
2
Q3  Q1 Q1  Q3
Coefficient of QD  Midhinge 
Q3  Q1 2
• It is not affected by extreme values.
• QD is based on central 50% of the observations.
• It can be calculated for the distribution having open-
ended classes.
The Interquartile Range (IQR)
• The IQR is Q3-Q1 and measures the spread in the middle
50% of the data
• The IQR is also called the midspread because it covers the
middle 50% of the data
• The IQR is a measure of variability that is not influenced by
outliers or extreme values
• Measures like Q1, Q3 and IQR that are not influenced by
outliers are called resistant measures
Mean Deviation (Average Deviation)
• Mean deviation or average deviation is the arithmetic mean
of the deviations of a given set of observations from their
average taking all the deviations as positive.
• Mean deviation shows the scatteredness around an
average.
• The mean deviation is also known as average absolute
deviation.
• It shows the average amount by which the items differ from
the mean or the median.
• While calculating average deviation we, have ignored the
fact that some deviation are positive and some are
negative. This is done because while measuring dispersion
we are interested in the amount and not the direction of
the variation.
For individual data set
1
Mean deviation from mean 
n  x x
1
Mean deviation from median 
n  x  Med
1
Mean deviation from mode 
n  x  Mode
For frequency distribution
1
Mean deviation from mean 
N  f x x
1
Mean deviation from median 
N  f x  Med
1
Mean deviation from mode 
N  f x  Mode
Mean Deviation (Average Deviation)
Merits
i) It is rigidly defined
ii) It is based on all observations of the series.
iii) It is simple to understand and easy to calculate.
iv) It is less affected by extreme observations.
Demerits
i) It ignores the negative sign.
ii) It cannot be computed in case of open end classes.
iii) It is not suitable for further mathematical treatment.
iv) It is affected by sampling fluctuations.
v) It does not give the satisfactory result when deviations are
taken from mode when mode is ill-defined.
Standard Deviation
• The standard deviation is the square root of the arithmetic
average of the squared deviation of a given set of
observations from their mean and is denoted by sigma   
• It is free from all the defects. Has the same units as the
original data
• Standard deviation is the best measure of dispersion and
most widely used in practice.
• Thus for a given set of observations, the formula for
calculating the standard deviation is given by
Population Standard Deviation
In case of individual data
1 12

N x    
N
x 2
 (  ) 2

In case of frequency distribution

1 2 1

N  f x    
N  fx 2  (  ) 2
Sample Standard Deviation
In case of individual data
1 2 1 n
s
n 1  x  x  
n  1
x 2

n 1
( x ) 2

In case of frequency distribution

1 2 1 N
s
N1  f x  x  
N1  fx 2 
N1
(x)2
Measures of Variation: The Sample Variance
• Average (approximately) of squared deviations of values
from the mean
• Sample variance:
n 2

 X i  X 
S 2  i 1
n 1
Where,

X arithmetic mean n sample size X i i valueof the variable X

th
Sample Standard Deviation: Calculation Example

Sample Data  X i  : 10 12 14 15 17 18 18 24
n=8 Mean  X 16

10  X   12  X   14  X     24  X 

2 2 2 2

S
n 1

10  16   12  16   14  16     24  16 

2 2 2 2


8 1
130 A measure of the “average”
 4.3095 scatter around the mean
7
Comparing Standard Deviations
Data A
Mean = 15.5
s = 3.338
Data B
Mean = 15.5
s = 0.926
Data C
Mean = 15.5
s = 4.567
Combined Standard Deviations

 N 1 1  N 2  2  N 1 d 1  N 2 d 2
2 2 2 2


12 N1  N 2
Where,

 12 combined standard deviation,

 1 standard deviation of first group.

 2 standard deviation of second group.


d 1  x1  x12 
d2 x 2  x12
x12 combined mean of series
Measures of Variation: Summary Characteristics
• The more the data are spread out, the greater the range,
variance, and standard deviation.
• The more the data are concentrated, the smaller the range,
variance, and standard deviation.
• If the values are all the same (no variation), all these
measures will be zero.
• None of these measures are ever negative.
The Coefficient of Variation
• Measures relative variation
• Always in percentage (%)
• Shows variation relative to mean
• Can be used to compare the variability of two or more sets
of data measured in different units

 S 
CV   100%
X
Comparing Coefficients of Variation
• Stock A:
• Mean price last year = $50
• Standard deviation = $5
S $5
CVA   100%  100% 10%
X $50
Both stocks have the
• Stock B: same standard
deviation, but stock B
• Mean price last year = $100 is less variable relative
• Standard deviation = $5 to its mean price

 S $5
CVB   100%  100% 5%
X $100
Comparing Coefficients of Variation
• Stock A:
• Mean price last year = $50
• Standard deviation = $5 Stock C has a
much smaller
S $5 standard
CVA   100%  100% 10%
X $50 deviation but a
much higher
• Stock C: coefficient of
variation
• Mean price last year = $8
• Standard deviation = $2
S $2
CVC   100%  100% 25%
X $8
Shape of a Distribution
• Describes how data are distributed
• Two useful shape related statistics are:
• Skewness
▪ Measures the extent to which data values are not
symmetrical
• Kurtosis
▪ Kurtosis measures the peakedness of the curve of
the distribution—that is, how sharply the curve rises
approaching the center of the distribution
Shape of a Distribution (Skewness)
• Measures the extent to which data is not symmetrical
Shape of a Distribution
Kurtosis Measures How Sharply the Curve Rises
Approaching the Center of the Distribution
Five Number Summary and the Boxplot
• The five numbers that help describe the center, spread and
shape of data are:
• Smallest observation,
• Largest Observation and
• Quartiles (Q1, Q2, and Q3)

Advantages of Box-and-Whisker Plot

1. Visually displays the center, spread, and the skewness of a
data set.
2. Clearly identifies outliers.
3. Helps to compare different distributions.
Five Number Summary: Shape of Boxplots
• If data are symmetric around the median then the box and
central line are centered between the endpoints
• A Boxplot can be shown in either a vertical or horizontal
orientation
How to Detect Outliers with Box–Plot
• A value that falls outside either of the inner fences is called
an outlier.
• An outlier could be:
 Mild or
 Extreme
• A mild outlier occurs when a value falls outside any of the
inner fences but inside either a lower or upper outer fence.
• An extreme outlier is a value that falls outside either of the
outer fences.
Inner Fences:
Lower inner fence = 1.5(IQR) below the Q1 = Q1 - 1.5(IQR)
Upper inner fence = 1.5(IQR) above the Q3 = Q3 + 1.5(IQR)
Outer Fences:
Lower outer fence = 3(IQR) below Q1 = Q1 - 3(IQR)
Upper outer fence = 3(IQR) above Q = Q + 3(IQR)
Relationships Among the Five-Number Summary and
Distribution Shape
In the data set of N = 90 ordered observations as shown below examine the
outliers:
30, 171, 184, 201, 212, 250, 265, 270, 272, 289, 305, 306, 322, 322, 336, 346, 351, 370, 390,
404, 409, 411, 436, 437, 439, 441, 444, 448, 451, 453, 470, 480, 482, 487, 494, 495, 499, 503,
514, 521, 522, 527, 548, 550, 559, 560, 570, 572, 574, 578, 585, 592, 592, 607, 616, 618, 621,
629, 637, 638, 640, 656, 668, 707, 709, 719, 737, 739, 752, 758, 766, 792, 792, 794, 802, 818,
830, 832, 843, 858, 860, 869, 918, 925, 953, 991, 1000, 1005, 1068, 1441
The computations are as follows:
• Median = (n+1)/2 largest data point = the average of the 45th and 46th ordered points
= (559 + 560)/2 = 559.5
• Q1 = .25(N+1)th ordered point = 22.75th ordered point = 411 + .75(436-411) = 429.75
• Q3 = .75(N+1)th ordered point = 68.25th ordered point = 739 +.25(752-739) = 742.25
• Interquartile range = Q3-Q1 = 742.25 - 429.75 = 312.5
• Lower inner fence = 429.75 - 1.5 (312.5) = -39.0
• Upper inner fence = 742.25 + 1.5 (312.5) = 1211.0
• Lower outer fence = 429.75 - 3.0 (312.5) = -507.75
• Upper outer fence = 742.25 + 3.0 (312.5) = 1679.75
From an examination of the fence points and the data, one point (1441) exceeds the
upper inner fence and stands out as a mild outlier; there are no extreme outliers.
Calculating The Interquartile Range
Example:
Distribution Shape and The Boxplot

Left-Skewed Symmetric Right-Skewed

Boxplot Example
• Below is a Boxplot for the following data:

• The data are right skewed, as the plot depicts

Numerical Descriptive Measures for a Population
• Descriptive statistics discussed previously described a
sample, not the population.
• Summary measures describing a population, called
parameters, are denoted with Greek letters.
• Important population parameters are the population
mean, variance, and standard deviation.
Numerical Descriptive Measures for a Population:
The mean µ
• The population mean is the sum of the values in the
population divided by the population size, N
N

X i
X1  X 2   X N
 i =1

N N
Where,
μ = population mean
N = population size
Xi = ith value of the variable X
Numerical Descriptive Measures for a Population:
The Variance Sigma Squared
• Average of squared deviations of values from the mean.
N

 X  
2
i
– Population variance:  2  i 1
N

Where,
μ = population mean
N = population size
Xi = ith value of the variable X
Numerical Descriptive Measures for a Population: The
Standard Deviation Sigma

• Most commonly used measure of variation

• Shows variation about the mean
• Is the square root of the population variance
• Has the same units as the original data
N

 X  
2
i

– Population standard deviation:  

i 1

N
Sample Statistics Versus Population Parameters
The Empirical Rule
• The empirical rule approximates the variation of data in a
bell-shaped distribution
• Approximately 68% of the data in a bell shaped distribution
is within 1 standard deviation of the mean or  1
Using the Empirical Rule
• Suppose that the variable Math SAT scores is bell-shaped
with a mean of 500 and a standard deviation of 90. Then,

– Approximately 68% of all test takers scored between

410 and 590, 500 90 .
– Approximately 95% of all test takers scored between
320 and 680, 500 180 .
– Approximately 99.7% of all test takers scored between
230 and 770, 500 270 .
Standard Deviation of the Combined Series
If n1, n2 are the sizes, X1, X2 are the means and σ1, σ2 are the
standard deviation of the series, then the standard deviation
of the combined series is

,,
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 are depicted here as under.
1. It is independent of change in the origin. This means that the value of the
standard deviation will remain the same, even if, each of the items of a series
is added or subtracted by a constant quantity.

Proof. Let there be 5 items in a series as 1,2,3,4 and 5. The standard deviations of these items as
such and after of these items as such and after a change in their origin will be as under.
 √ ( )
∑𝑥
2
1
𝜎=
𝑛
∑ 2
𝑥 −
𝑛

Thus, It is proved that the in all the cases of change of the origin
remains the same.
2. It is dependent of the change in the scale. This means that the
value of standard deviation is multiplied, or divided by the same
contact by which each item of the series is multiplied, or divided.
Proof. Let there be 5 items in a series as 5,10,15,20 and 25. The
standard deviation of these items as such, and after the change
in their scale will be as under.
Measures of Skewness
Various measures of skewness (Sk) are :
(1) Sk = Mean – Median = M – Md or (Pearsons’ Measure)
Sk = Mean – Mode = M – Mo
(2) Sk = (Q3 – Md) – (Md – Q1) = Q3 + Q1 – 2 Md (Bowley’s Measure)

(3) Sk = (P90 – P50) – (P50 – P10) = P90 + P10 – 2P50 (Kelly’s Measure)

These are the absolute measures of skewness and are not of much
practical utility because of the following reasons :
(i) Since the absolute measures of skewness involve the units of
measurement, they cannot be used for comparative study of the
two distributions measured in different units of measurement.
(ii) Even if the distributions are having the same units of measurement,
the absolute measures are not recommended because we may
come across different distributions which have more or less
identical skewness (absolute measures) but which vary widely in the
• Thus for comparing two or more distributions for skewness we
compute the relative measures of skewness, also commonly known
as coefficients of skewness which are pure numbers independent of
the units of measurement.
• Moreover, in a relative measure of skewness, the disturbing factor of
variation or dispersion is eliminated by dividing the absolute measure
of skewness by a suitable measure of dispersion.
• The following are the coefficients of skewness which are commonly
used.
Karl Pearson’s Coefficient of Skewness.
• This is given by the formula : Sk = (Mean – Mode)/ SD = (M – Mo)/σ
• But quite often, mode is ill-defined and is thus quite difficult to
locate. In such a situation, we use the following empirical relationship
between the mean, median and mode for a moderately
asymmetrical (skewed) distribution : Mo = 3Md – 2M
• Substituting in above, we get Sk = M – (3Md – 2M) / σ = 3(M – Md)/σ
1. Theoretically, Karl Pearson’s Coefficient of Skewness lies between the limits
± 3, but these limits are rarely attained in practice
2. Skewness is zero if M = Mo = Md.
• In other words, for a symmetrical distribution mean, mode and median
coincide i.e., M = Md = Mo.
3. Sk > 0, if M > Md > Mo or if Mo < Md < M
• Thus, for a positively skewed distribution, the value of the mean is the
greatest of the three measures and the value of mode is the least of the
three measures.
• If the distribution is negatively skewed, then the inequality
Sk < 0, if M < Md < Mo or if Mo > Md > M
In other words, for a negatively skewed distribution, of the three
measures of central tendency viz., mean, median and mode, the mode
has the maximum value and the mean has the least value.
4. While ‘dispersion’ studies the degree of variation in the given distribution.
Skewness attempts at studying the direction of variation. Extreme variations
towards higher values of the variable give a positively skewed distribution
while in a negatively skewed distribution, the extreme variations are towards
the lower values of the variable.
5. In Pearson’s coefficient of skewness, the disturbing factor of
variation is eliminated by dividing the absolute measure of skewness
M – Mo by the measure of dispersion σ (standard deviation).

6. If the distribution is symmetrical, say, about mean, then

Mean – Xmin = Xmax – Mean
⇒ Xmax + Xmin = 2 Mean
Also Xmax – Xmin = Range

Adding and subtracting two we get respectively :

2 Xmax = 2 Mean + Range ⇒ Xmax = Mean + 1 /2*Range

2Xmin = 2Mean – Range ⇒ Xmin = Mean - 1 /2*Range

Example
In a symmetrical distribution the mean, standard deviation and range of marks
for a group of 20 students are 50, 10 and 30 respectively. Find the mean marks
and standard deviation of marks, if the students with the highest and the
lowest marks are excluded.
Bowley’s Coefficient of Skewness.
Bowley’s coefficient of skewness is based on the quartiles and is given
by

Remarks
1. Bowley’s coefficient of skewness is also known as Quartile
coefficient of skewness and is especially useful in situations where
quartiles and median are used :
(i) When the mode is ill-defined and extreme observations are
present in the data.
(ii) When the distribution has open end classes or unequal class
intervals. In these situations, Pearson’s coefficient of skewness
cannot be used.
2. From (above), we observe that: Sk = 0, if Q3 – Md = Md – Q1
This implies that for a symmetrical distribution (S k = 0), median is
equidistant from the upper and lower quartiles.
Moreover, skewness is positive if :
Skewness is Positive if:

Skewness is Negative if:

3. Limits for Bowley’s Coefficient of Skewness.

Bowley’s coefficient of skewness ranges from –1 to 1.

Further

4. It should be clearly understood that the values of the coefficient of

skewness obtained by Bowley’s formula and Pearson’s formula are
not comparable, although in each case, Sk = 0 implies the absence of
skewness i.e., the distribution is symmetrical. It may even happen
that one of them gives positive skewness while the other gives
negative skewness.
Kelly’s Measure of Skewness
• The drawback of Bowley’s coefficient of skewness, (that it ignores
the 50% of the data towards the extremes), can be partially
removed by taking two deciles or percentiles equidistant from the
median value.
• The refinement was suggested by Kelly. Kelly’s percentile (or decile)
measure of skewness is given by:
Sk = (P90 – P50) – (P50 – P10) = P90 + P10 – 2P50
But P90 = D9 and P10 = D1. Hence, can be re-written as:
Sk = (D9 – D5) – (D5 – D1) = D9 + D1 – 2D5
Pi and Di are the ith percentile and decile respectively of the given
distribution.
• However, for practical purposes, we generally compute the
coefficient of skewness, which is given by:
Sk (Kelly) = (P90 – P50) – (P50 – P10)/ (P90 – P50) + (P50 – P10)
= P90 + P10 – 2P50 / (P90 –P10) OR
Remarks 1.
We have:
D5 = P50 = Median. Hence, we get
Sk (Kelly) = P90 + P10 – 2Md /P90 – P10 = D9 + D1 – 2Md/ D9 – D1
This method is primarily of theoretical importance only and is seldom
used in practice.
In statistics, moments are quantitative measures that capture various
aspects of the shape and characteristics of a distribution. They are
derived from the powers of the deviations of the data points from a
specific point, typically the mean. Moments help describe the central
tendency, variability, skewness, and kurtosis of a dataset.
Types of Moments
1. Raw Moments (About Zero): These are calculated about zero and
are given by:

where xi is each data point, k is the order of the moment, and n is the
number of data points.
2. Central Moments (About the Mean): These are calculated about the
mean (μ) and are given by:

Central moments provide insight into the distribution's shape.

Key Moments

1. First Moment:
The mean (μ), which represents the central location of the data.

2. Second Moment:
The variance (σ2), which measures the spread or dispersion of the
data. It is the average squared deviation from the mean.

3. Third Moment:
Related to skewness, which indicates the asymmetry of the
distribution. A positive skew suggests a longer tail on the right, while
a negative skew indicates a longer tail on the left.

4. Fourth Moment:
Related to kurtosis, which measures the "tailedness" or the peak of
the distribution. High kurtosis indicates heavy tails, while low
kurtosis suggests lighter tails.
Importance of Moments

• Descriptive Statistics: Moments provide a systematic way to

describe and summarize distributions.
• Modeling: Moments are used in defining probability distributions
and fitting models to data.
• Statistical Inference: Higher-order moments are often used to derive
properties of estimators and tests.
• Moments are fundamental in both descriptive and inferential
statistics, forming the basis for understanding and interpreting data
distributions.
Moments are powerful statistical tools that help describe key
characteristics of a dataset, such as central tendency, variability,
skewness, and kurtosis. Here’s how they work in detail:
1. Central Tendency (1st Moment)
• The first moment about zero is the mean of the dataset, which
measures the central location or average value.
• Formula:

Interpretation: The mean tells us where the center of the data lies.
2. Variability (2nd Moment)
• The second central moment is the variance, which measures the
spread or dispersion of the data around the mean.
• Formula:

• The square root of variance is the standard deviation, which

provides a measure of dispersion in the same units as the data

Interpretation:
• A larger variance indicates that the data points are spread out more
widely from the mean, while a smaller variance indicates they are
closer to the mean.
3. Skewness (3rd Moment)
• The third central moment describes the skewness of the dataset,
which measures the asymmetry of the distribution.
• Formula (Standardized Skewness):

Interpretation:
• Positive Skewness: The right tail is longer, and the data is skewed to
the right.
• Negative Skewness: The left tail is longer, and the data is skewed to
the left.
• Zero Skewness: The distribution is symmetric.
4. Kurtosis (4th Moment)
• The fourth central moment relates to kurtosis, which measures the
"tailedness" or the peakiness of the distribution.
• Formula (Excess Kurtosis):

• (Subtracting 3 makes the kurtosis of a normal distribution

equal to 0.)

Interpretation:
• High Kurtosis (>0): Indicates heavy tails and sharp peaks
(leptokurtic distributions).
• Low Kurtosis (<0): Indicates light tails and flatter peaks
(platykurtic distributions).
• Normal Kurtosis (≈0): Indicates a mesokurtic distribution like
the normal distribution
Summary of How Moments Describe a Dataset
1. 1st Moment (Mean): Central location.
2. 2nd Moment (Variance): Spread or variability.
3. 3rd Moment (Skewness): Asymmetry or direction of the tail.
4. 4th Moment (Kurtosis): Tailedness or shape of the peak.

These moments together provide a comprehensive understanding of the

data's shape and characteristics, enabling deeper insights into its
distribution and behavior.
Chebyshev Rule
 1 
• Regardless of how the data are distributed, at least  1  2 
100%
 k 
of the values will fall within k standard deviations of the mean
(for k > 1).

– Examples: