Scribd
Upload
36 views22 pages

Variance and Coefficient of Variation

The document discusses various measures of variation (dispersion) in data, including population variance, sample variance, standard deviation, coefficient of variation, interquartile range, and box plots. It provides formulas for calculating these measures and includes examples to demonstrate calculating variance, standard deviation, quartiles, and constructing a box plot from a data set.

Uploaded by

Hatem Dheer
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
36 views22 pages

Variance and Coefficient of Variation

The document discusses various measures of variation (dispersion) in data, including population variance, sample variance, standard deviation, coefficient of variation, interquartile range, and box plots. It provides formulas for calculating these measures and includes examples to demonstrate calculating variance, standard deviation, quartiles, and constructing a box plot from a data set.

Uploaded by

Hatem Dheer
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd

Measures of variation

(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
Formula:
1. For ungrouped data population variance:
N

 (x i  )2
 2
 i 1

2. For ungrouped data sample variance


n

2
 (x
i 1
i  x) 2

S 
n 1
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
Standard Deviation (S)
1. It is the square root of variance.
2. Most commonly used measure of
variance.
3. Shows variation about mean.
4. It used to compare between more than
one data set when the means are equal, the
best one is the minimum.

S  S2
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Example
 2
Year of No. of   ( x  x )  ( x  x )
graduation Students
2004 4 -2 4
2005 6 0 0
2006 5 -1 1
2007 8 2 4
2008 7 1 1
Total 30 0 10

• Calculate variance and standard deviation


Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
n

 (x i  x )2
10
• Solution: Variance 2
S  i 1
n 1

4
 2.5

• Standard Deviation S  S 2  2.5  1.58

• Interpretation: The observations fall 1.58


units from the mean.
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
Variance and standard deviation for grouped
data

• Formula
n

n ( f x )
i i
2

2
fx
i 1
i
2
i  i 1

n
S 
n 1
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Example: the table below shows the temperature of a sample of
50 cities taken at the same time on a certain day; determine the
mean and standard deviation of the sample.
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬

Temp. f Cum Midp x2 f .x f .x 2


ulativ oint
e F.
10-14 10 10 12 144 120 1440
15-19 12 22 17 289 204 3468
20-24 18 40 22 484 396 8712
25-29 6 46 27 729 162 4374
30-34 4 50 32 1024 128 4096
Total 50       1010 22090
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬

x
 f .x 1010
  20.2
f 50
n

n  i i
( f x ) 2

(1010) 2
2
fx
i 1
i
2
i  i 1

n
22090 
50 1688
S     34.45
n 1 49 49
2
S  S  34.45  5.87
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Coefficient of Variation (C.V) ‫ختالف‬
( ‫)مع امل ا ال‬
1. It is the main important application of the
mean and standard deviation.
2. Measures relative variation and always in
percentage (%).
3. Can be used to compare two or more data
sets measured in different units.
4. Can be used widely in chemistry and
engineering science.
5. The variable with smaller C.V is less
dispersed than others so it is the better.
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Formula:
S
• Coefficient of Variation C.V   100%
x

• Example: Suppose that technician A


completes 40 analysis daily with standard
deviation of 5, technician B completes 160
analysis per day with standard deviation of
15.
• Which employee shows less variability or
better?
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Sol.
S 5
C.V ( A)   100%   100%  12.5%
x 40
S 15
C.V ( B)   100%   100%  9.4%
x 160

• Employee B is better than A because


he have the less variation.
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
Interquartile Range [IQR] (‫)ا لمدىا لربيعي‬
There are three quartiles Q1, Q2, Q3
1. Q1 is a 25% of sorted data.
2. Q2 is a 50% of sorted data or median.
3. Q3 is a 75% of sorted data.
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
Formulas
 N 
  F1 
Q1  L1   4   C1
 f 1 

 

Q2 the same as median formula.

 3N 
 4  F3 
Q3  L3     C3
 f3 
 
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Example: you have the frequency table:
Class boundaries Frequency (f) Cumulative frequency (F)
31.5 – 36.5 4 4
36.5 – 41.5 7 11
41.5 – 46.5 10 21
46.5 – 51.5 7 28
51.5 – 56.5 18 46
65.5 – 61.5 4 50

Calculate Q1, Q2, Q3 and interquartile range.


Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
 N 

• Sol.   F1 
Q1  L1   4   C1
 f1 

 

50
step 2   12.5
4

Step (3) the first quartile class is [41.5-46.5]

step(4) : L1  41.5, F1  11, f1  10, C1  5

N   50 
4  F1  4  11 
Q1  L1    1 C  41 .5     5  42.25
 f 1   10 
   
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Sol. Q2
50
step 2   25
2
Step (3) the median quartile class is [46.5-51.5]

step(4) : L1  46.5, F1  21, f 1  7, C1  5

N 
2  F2
 25  21
Q2  L    2 C  46 .5     5  49.36
 f2   7 
 
Measures of variation
(dispersion) ‫ت‬
[ ‫]مقاييسا لتشت‬
• Sol. Q3
50  3
step 2   37.5
4
Step (3) the third quartile class is [51.5-56.5]
step(4) : L3  51.5, F3  28, f 3  18, C 3  5

 N 3   3  50 
  F3   4  28 
Q3  L3   4   C 3  51.5     5  54.14
 f 3   18 
   

Interquartile range = Q3 - Q1 =54.14-42.25=11.89.


Box plot
• A box plot is a descriptive
statistics and it is a convenient
way of graphically depicting
groups of numerical data through
their quartiles.
• Example: Plot a box plot for {7, 4,
3, 5, 6, 8, 10, 1}
Box plot
• Solution:
• Sort data as: {1, 3, 4, 5, 6, 7, 8, 10}.
• Minimum value is 1, maximum value is
10.
• Calculate Q1, Q2, Q3 as:
k (n  1) 25(9)
position of Q1    2.25
100 100

3 4
Q1   3 .5
2
Box plot
• Q2 and Q3
k (n  1) 50(9)
position of Q 2    4.5
100 100
56
Q2   5.5
2
k (n  1) 75(9)
position of Q3    6.75
100 100
78
Q3   7.5
2
Box plot
Box plot
Note: The above plotting done by computer using
software R as:

> A <- c(7, 4, 3, 5, 6, 8, 10 ,1)


> quantile(A)

0% 25% 50% 75% 100%


1.00 3.75 5.50 7.25 10.00

You might also like