You are on page 1of 14

1

Unit Four
Measure of Variation

seid.belay@aastu.edu.et 11/26/2023
2 Objectives

 To describe dispersion /variability/ in a data.


 To compare the spread in two or more distributions.
 To determine the reliability of an average.

seid.belay@aastu.edu.et 11/26/2023
3 Types of Measures of Variation
 Absolute Measures of Variation:
 Has the same measurement as the data,
 Used for comparing variability of datasets same measurement units
 Examples: Range, standard deviation, mean deviation
 Relative Measures of variation:
 Unit less,
 Used for comparing variability of datasets varying in measurement units
 Examples: Relative Range, Coefficient of variation, coefficient of mean deviation

seid.belay@aastu.edu.et 11/26/2023
4 Type of Measures variation
  Affected by extreme values.
Range
 It does not take into account all observations.
 It is easy to calculate and simple to understand.
 Relative Range =  It does not tell anything about the distribution of values in
the set of data relative to some measures of central
tendency.
 Inter quartile Range (IQR) = Q3 – Q1
 It is relatively simple to understand
 Mean Absolute Deviation:  Its computation is simple.
 It is less affected by extreme values
 Coefficient of mean Deviation:  It is better than the range and relative
range since it is based on all
observations.
 It is not suitable for further statistical
treatment.

seid.belay@aastu.edu.et 11/26/2023
5 Variance, Standard Deviation, cv

(∑ 𝐱 )
𝐍 𝟐
Variance: Average squared deviation from mean 𝐢
𝐍 𝐍 𝐍
𝐢 =𝟏
∑ ( 𝐱𝐢 −𝛍 ) 𝟐
∑ 𝐱𝐢 − 𝟐
𝐍 ∑ 𝐱𝟐
𝐢
𝛔𝟐 = 𝐢 =𝟏 = 𝐢 =𝟏 = 𝐢 =𝟏
− 𝛍𝟐
 Population Variance: 𝐍 𝐍 𝐍
(∑ 𝐱 )
𝐧 2

𝐧 𝐧 𝐢 𝐧
𝐢=1
∑ ( 𝐱𝐢− 𝐱) 2
∑ 𝐱𝐢 − 2
𝐧 ∑ 𝐱 2𝐢 − 𝐧 𝐱 2
𝐒 2= 𝐢 = 1 = 𝐢=1
= 𝐢=1
 Sample Variance: 𝐧− 1 𝐧 −1 𝐧 −1

 Standard Deviation: principal square root of variance


 Used to measure the spread of data, consistency of a variable and the proportion of data values
that fall within a specified interval in a distribution
 for population, for sample

 Coefficient of variation: the ratio standard deviation to the mean


 Used to compare variability of two or more datasets
seid.belay@aastu.edu.et 11/26/2023
6 Example
X 1 2 3 4 5 6 7
 Given sample data:
f 5 10 9 6 3 2 1

 Compute: Range, relative range, Mean deviation, coefficient of mean deviation, variance, standard
deviation and cv x f xf d |d| f|d| dd fdd
1 5 5 -2.1 2.1 10.5 4.41 22.05
 Range = 7-1 = 6 2 10 20 -1.1 1.1 11 1.21 12.1
 Relative range = 6/8 3 9 27 -0.1 0.1 0.9 0.01 0.09
4 6 24 0.9 0.9 5.4 0.81 4.86
 Mean =
5 3 15 1.9 1.9 5.7 3.61 10.83
 Mean Deviation = 2 6 2 12 2.9 2.9 5.8 8.41 16.82
 7 1 7 3.9 3.9 3.9 15.21 15.21
Coefficient of MD = 1.2/3.1 = 0.387
T 36 110 43.2 81.96
 Variance = = 81.96/35 =2.34 Av 3.1 1.2 2.34
 Standard Deviation = s= =1.53

seid.belay@aastu.edu.et 11/26/2023
7 Example
1. The mean weight of 20 children was found to be 30 kg with variance of 16kg 2 and their mean height was
150 cm with variance of 25cm2. Compare the variability of weight and height of these children.

2. For the frequency distribution below, Compute the variance, standard deviation and cv

cm t fcm d dd fdd
Class Frequency
2.5 8 20 -5.6 31.36 250.88
0-5 8 7.5 9 67.5 -0.6 0.36 3.24
12.5 4 50 4.4 19.36 77.44
5-10 10 17.5 2 35 9.4 88.36 176.72
10-15 3 22.5 1 22.5 14.4 207.36 207.36
Sum 24 195 715.64
15-20 2 8.1 31.11
20-25 1 5.58

seid.belay@aastu.edu.et 11/26/2023
8 Properties of Variance

Example:
The mean temperature in AA is 77 degrees Fahrenheit with standard deviation 9 degrees Fahrenheit. What is the
mean temperature in degrees Celsius? What is the standard deviation in degrees Celsius?
Sol. Let X be temperature in degree Fahrenheit . E(X)= = 77 and 9 and
Y be temperature in degree Celsius, Hence, .
Then
 E(Y) = 5/9(E(X) -32)=(5/9)(45)=25 and (Properties of mean)
 = (5/9)(5/9) = (25/81)*81 =25. (Properties of variance)
seid.belay@aastu.edu.et 11/26/2023
Therefore, standard deviation of Y = 5.
9 Range Rule of Thumb and Chebyshev’s Theorem
 Range Rule of Thumb
 Let R be the range of a data set.
 Rough Estimate of standard deviation is given by
 If standard deviation S is known, Rough Estimate of the minimum and
maximum values are:

 Maximum usual value


 Chebyshev’s Theorem
 The proportion or fraction of any set of data lying within k standard Or
deviation of the mean is always at least
 If k=2, ¾ or at least 75% observations lie within 2 standard deviations

 Example: A data set has n = 900, = 45 and s = 3. At least how many data points
lie between 40 and 50? (64%n = 576)

seid.belay@aastu.edu.et 11/26/2023
10 Standard Scores
 Describes the relative position of an observation in the entire distribution in terms of the mean and
standard deviation. For i-th observation Note:
Mean of z score =0
 Variance of z score =1
Interpretation:
 A student scored 65 on a calculus test that had a mean of 50 and a standard deviation of 10; she
scored 30 on a algebra test with a mean of 25 and a standard deviation of 5. Compare her relative
positions on each test.
 Sol:

Algebra Calculus Z_Score Relative to the class, she performed better in


Mean 25 50 1 calculus, which is 1 and half standard
S.d 5 10 1.5 deviation above the average
Score 30 65

seid.belay@aastu.edu.et 11/26/2023
11 Moment
 Definition:
A Moment is average of deviations from an arbitrary origin raised to an integral power
of the observations of a distribution. i.e. , the r-th moment is defined as:
 r-th Central Moment/Moment about the mean/:

 Special cases: 0, and


 r-th Moment about the origin/Moment about 0/:

 Special cases: 1, and

seid.belay@aastu.edu.et 11/26/2023
12 Skewness
 Measure of lack of symmetry of a distribution
 Pearson Coefficient of Skewness:

 For moderately
 Interpretation:
 Example: If the mean, mode and standard deviation of a frequency
distribution are 70.2, 73.6, and 6.4, respectively. What can you
state about its skewness?

 Negatively skewed
 Large values are more frequent than smaller values

seid.belay@aastu.edu.et 11/26/2023
13 Kurtosis
 Measures the degree of Preakness of a symmetric distribution
 Highly peaked – leptokurtic
 Moderately peaked – Mesokurtic
 Flat- topped – platykurtic
 Moment coefficient of kurtosis is Defined as =

If

seid.belay@aastu.edu.et 11/26/2023
14

Thank You All!

seid.belay@aastu.edu.et 11/26/2023

You might also like