Measures of Skewness and Kurtosis

MEASURES OF
SKEWNESS AND
KURTOSIS
SKEWNESS
Symmetric or Asymmetric
SKEWNESS
■ The degree of asymmetry of a distribution

■ Indicates not only the amount of asymmetry but also the
direction of the distribution
■ The greater the value of the skewness to 0, the nearer the

distribution is to a normal distribution.
3
SYMMETRIC or ASYMMETRIC
■ When the mean, median and mode have identical values, this
suggests that the distribution is symmetric.
■ If the mean, median and mode varies, than the distribution is

characterized as asymmetric or skewed.
4
Two Types of
Skewness
Positive or Negative
Positive Skewness
■ Refers to the distribution wherein the longer tail is directed to the

right or to the bottom if the data is presented sideways
■ In this kind, more observations are concentrated on the left of the

figure.
■ Sometimes called as “skewed to the right”
6
Positive Skewness
7
Negative Skewness
■ Refers to the distribution wherein the longer tail is directed to the left
■ In this, more observations are concentrated on the right of the figure.
■ Sometimes called as “skewed to the left”
■ In either case, the median always lies between the mean and the
mode
○ Right skewed = mean > median
○ Left skewed = mean < median
8
Skewness
■ Positive value of coeffficient = skewed to the right

■ Negative value of coefficient = skewed to the left
■ Zero value of coefficient = normal distribution
9
Example:
■ The ages of 10 patients entering the general hospital are as

follows:
15, 31, 75, 84, 19, 79, 74, 78, 76, 29
Determine and interpret the moment coefficient of skewness.
10
Steps in Determining the Coefficient of
Skewness (raw data):
1. Compute the sample mean and standard deviation.
2. Subtract the mean from the individual observations to get
the deviation.
3. Take the cube of the deviation and then get the sum
4. Get the cube of the sample standard deviation then multiply
to (n-1)
5. Divide the answer in step 3 by the answer in step 4. The
quotient is the value of the moment coefficient of skewness.
11
Steps in Determining the Coefficient of Skewness
(grouped data):
1. Compute the sample mean and standard deviation.
2. Subtract the mean from the midpoint of each class interval to
get the deviation.
3. Multiply the frequency of each class interval to its
corresponding cube of the deviation and then get their sum.
4. Get the cube of the sample standard deviation and then
multiply by (n-1)
quotient is the value of the moment coefficient of skewness.
12
KURTOSIS
Leptokurtic, Mesokurtic,
Platykurtic
KURTOSIS
14
KURTOSIS
■ A measure of the degree of peakedness or flatness relative to

a normal distribution
■ The concern is in the height of the curve.
■ Data set with high kurtosis tend to have a distinct peak near
the mean, decline rather rapidly, and have heavy tails.
15
LEPTOKURTIC
■ A distribution having a relatively high peak

■ lepto (slender)
16
MESOKURTIC
■ Refers to the distribution neither very peaked nor very flat-

topped
17
PLATYKURTIC
■ The “normal” or ideal distribution having relatively flat-top

■ “platy” = flat or broad
18
Formula for Coefficient of Kurtosis
(raw data)
■ If KU > 3, then distribution is leptokurtic or more peaked than the

normal curve;
■ If KU = 3, the distribution is mesokurtic; and
■ If KU < 3, the distribution is platykurtic or less peaked than the
normal curve.
19
Steps in Determining the KURTOSIS
(raw data)
1. Compute the mean and standard deviation.
2. Subtract the mean from the individual observations to get the
deviation,
3. Raise the deviation to the fourth power and then obtain the
sum.
4. Raise the standard deviation to the fourth power and then
multiply by (n-1)
quotient is the moment coefficient of kurtosis.
20
Steps in Determining the KURTOSIS
(grouped data)
1. Compute the mean and standard deviation.
2. Subtract the mean from the midpoint of each class interval to
get the deviation
3. Raise the deviation to the fourth power.
4. Multiply the frequency of each class interval to each
corresponding fourth power of the deviation to get values for
fi (X-Xi)4 column.
5. Find the sum of step 4 to get the summation.
6. Raise the standard deviation to the fourth power and then
multiply by (n-1).
21
PRACTICE
EXERCISES
Coefficient of Kurtosis
1.
Consider the data on birth weight (in

ounces) of the newly born children at
hospital X if the data are as follows:
112, 111, 107, 119, 92, 80, 81, 84, 118,
106, 103, and 94, compute the
coefficient of skewness and coefficient
of kurtosis.
23
2.
The distribution of age (in AGE f

years) of 45 patients 80-89 2
admitted at Philippine 70-79 4
General Hospital is 60-69 3
shown below. Calculate 50-59 10
the coefficient of 40-49 13
skewness and 30-39 8
coefficient of kurtosis. 20-29 4
10-19 1
24

Measures of Skewness and Kurtosis

Uploaded by

Copyright:

Available Formats

Measures of Skewness and Kurtosis

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Measures of Skewness and Kurtosis

Uploaded by

Copyright:

Available Formats

MEASURES OF

■ The degree of asymmetry of a distribution

■ The greater the value of the skewness to 0, the nearer the

■ If the mean, median and mode varies, than the distribution is

■ Refers to the distribution wherein the longer tail is directed to the

■ In this kind, more observations are concentrated on the left of the

■ Sometimes called as “skewed to the right”

■ Positive value of coeffficient = skewed to the right

■ The ages of 10 patients entering the general hospital are as

■ A measure of the degree of peakedness or flatness relative to

■ The concern is in the height of the curve.

■ A distribution having a relatively high peak

■ Refers to the distribution neither very peaked nor very flat-

■ The “normal” or ideal distribution having relatively flat-top

■ If KU > 3, then distribution is leptokurtic or more peaked than the

Consider the data on birth weight (in

The distribution of age (in AGE f

You might also like