Probability and statistics for Computer Science

CHAPTER 3
3. Measures of Central Tendency
3.1. Measures of Central Tendency
The Summation Notation:
- Let X1, X2, X3 …XN be a number of measurements where N is the total number of observation
and Xi is ith observation.
- Very often in statistics an algebraic expression of the form X1+X2+X3+...+XN is used in a
formula to compute a statistic. It is tedious to write an expression like this very often, so
mathematicians have developed a shorthand notation to represent a sum of scores, called the
summation notation.
- The symbol ∑ is the Mathematical shorthand representation of x1+ x2+ x3+…..+xN

PROPERTIES OF SUMMATION
n
1.  k  nk
i 1
where k is any constant
n n
2.  kX i  k X i where k is any constant
i 1 i 1
n n
3.  ( a  bX i )  na  b  X i where a and b are any constant
i 1 i 1
n n n
4. (X
i 1
i  Yi )   X i   Yi
i 1 i 1

Example: considering the following data determine

X Y
5 6
7 7
7 8
6 7
8 8
5 5
a)  X i = 33
i 1
e) (X i 1
i  Yi )  -3
5 5
b)  Yi
i 1
= 36 f) X Y
i 1
i i  241

1|Page
By Habitamu W. (MSc)
Probability and statistics for Computer Science

5 5

 10 =50. X  223
2
c) g) i
i 1 i 1
5 5 5
d)  ( X i  Yi )  69
i 1
h) ( X i )( Yi )  1188
i 1 i 1

Types of measures of central tendency

 The Mean (Arithmetic, Geometric and Harmonic)
 The Mode
 The Median
 Quantiles (Quartiles, Deciles and Percentiles)

The Arithmetic Mean

 given by:
X 1  X 2  ...  X n
X 
n
n

X i
X  i 1

n
 If X1 occurs f1 times, if X2occurs f2 times, … , if Xn occurs fn times
k

fX k

f
i i
Then the mean will be X  i 1
k
, where k is the number of classes and i n
f
i 1
i
i 1

Example: Obtain the mean of the following number

2, 7, 8, 2, 7, 3, 7
Solution:
Xi fi Xifi
2 2 4
3 1 3
7 3 21
8 1 8
Total 7 36

f i Xi
36
X  i 1
4
  5.15
f
7
i
i 1

Arithmetic Mean for Grouped Data

k

f i Xi th th
X  i 1
k
, Where Xi =the class mark of the i class and fi = the frequency of the i class
f i 1
i

2|Page
By Habitamu W. (MSc)
Probability and statistics for Computer Science

Example: calculate the mean for the following age distribution.

Class Frequency
6- 10 35
11- 15 23
16- 20 15
21- 25 12
26- 30 9
31- 35 6
Solutions:
 First find the class marks
 Find the product of frequency and class marks
 Find mean using the formula.

Class fi Xi Xifi
6- 10 35 8 280
11- 15 23 13 299
16- 20 15 18 270
21- 25 12 23 276
26- 30 9 28 252
31- 35 6 33 198
Total 100 1575

f i Xi
1575
X  i 1
6
  15.75
f
100
i
i 1

Special properties of Arithmetic mean

n
1. (X
i 1
i  X )  0.
n n

2.  ( Xi  X ) 2   ( X i  A) 2 , A  X
i 1 i 1

3. Let X 1 is the mean of n1 observations, X 2 is the mean of n2… and X k is the mean of nk
observations Then the mean of all the observation in all groups often called the combined
mean is given by:
k

X1n1  X 2 n 2  ....  X k n k 
Xini
Xc   i1k
n1  n 2  ...n k
n
i 1
i

Example: In a class there are 30 females and 70 males. If females averaged 60 in an

examination and boys averaged 72, find the mean for the entire class.
Solutions:

3|Page
By Habitamu W. (MSc)
Probability and statistics for Computer Science

Females Males
X 1  60 X 2  72
n1  30 n2  70

X n  X 2 n2 X n i i
Xc  1 1  i 1
n1  n2 2

n i 1
i

30(60)  70(72) 6840

 Xc    68.40
30  70 100

4. If a wrong figure has been used when calculating the mean the correct mean can be
obtained without repeating the whole process using:
(CorrectValue  WrongValue)
CorrectMean  WrongMean 
n
Where n is total number of observations.

Example: An average weight of 10 students was calculated to be 65.Latter it was discovered

that one weight was misread as 40 instead of 80 kg. Calculate the correct average weight.

Solutions:
(CorrectValue  WrongValue)
CorrectMean  WrongMean 
n
(80  40)
CorrectMean  65   65  4  69k.g.
10
5. The effect of transforming original series on the mean.
a) If a constant k is added/ subtracted to/from every observation then the new mean
will be the old mean± k respectively.
b) If every observations are multiplied by a constant k then the new mean will be k*old
mean
Example:
1. The mean of n Tetracycline Capsules X1, X2, …,Xn are known to be 12 gm. New set of
capsules of another drug are obtained by the linear transformation Yi = 2Xi – 0.5 ( i =
1, 2, …, n ) then what will be the mean of the new set of capsules
Solutions:
NewMean  2 * OldMean  0.5  2 * 12  0.5  23.5

2. The mean of a set of numbers is 500.

a) If 10 is added to each of the numbers in the set, then what will be the mean of the new
set?
b) If each of the numbers in the set are multiplied by -5, then what will be the mean of
the new set?
Solutions:
4|Page
By Habitamu W. (MSc)
 Probability and statistics for Computer Science

a).NewMean  OldMean  10  500  10  510

b).NewMean  5 * OldMean  5 * 500  2500

Weighted Mean
n

X W i i
Xw  i 1
n

W
i 1
i

Example:
A student obtained the following percentage in an examination:
English 60, Biology 75, Mathematics 63, Physics 59, and chemistry 55.Find the students
weighted arithmetic mean if weights 1, 2, 1, 3, 3 respectively are allotted to the subjects.
Solutions:
5

X W i i
60 * 1  75 * 2  63 * 1  59 * 3  55 * 3 615
Xw  i 1
   61.5
1 2  1 3  3
5
10
W
i 1
i

Merits and Demerits of Arithmetic Mean

Merits:
It is based on all observation.
It is suitable for further mathematical treatment.
It is stable average, i.e. it is not affected by fluctuations of sampling to some extent.
It is easy to calculate and simple to understand.
Demerits:
 It is affected by extreme observations.
 It cannot be used in the case of open end classes.
 It cannot be determined by the method of inspection.
 It cannot be used when dealing with qualitative characteristics
The Geometric Mean
 The geometric mean of a set of n observation is the nth root of their product.
 The geometric mean of X1, X2 ,X3 …Xn is denoted by G.M and given by:

G.M  n X 1 * X 2 * ... * X n
 Taking the logarithms of both sides
1
log(G.M)  log(n X 1 * X 2 * ... * X n )  log(X 1 * X 2 * ... * X n ) n
1 1
 log(G.M)  log(X 1 * X 2 * .... * X n )  (log X 1  log X 2  ...  log X n )
n n
n
1
 log(G.M)   log X i
n i 1
 The logarithm of the G.M of a set of observation is the arithmetic mean of
5|Page
By Habitamu W. (MSc)
 Probability and statistics for Computer Science

their logarithm.
1 n
 G.M  Anti log(  log X i )
n i 1
Example:
Find the G.M of the numbers 2, 4, 8.
Solutions:
G.M  n X1 * X2 * ... * Xn  3 2 * 4 * 8  3 64  4
Remark: The Geometric Mean is useful and appropriate for finding averages of
Ratios.

The Harmonic Mean

The harmonic mean of X1, X2 , X3 …Xn is denoted by H.M and given by:
n
H.M  n , This is called simple harmonic mean.
1

i 1 X i
In a case of frequency distribution:

k
n
H.M  k , n   fi
fi

i 1 X i
i 1

If observations X1, X2, …Xn have weights W1, W2, …Wn respectively, then their harmonic mean is given
by

W i
H.M  n
i 1
, This is called Weighted Harmonic Mean.
W
i 1
i Xi

Remark: The Harmonic Mean is useful and appropriate in finding average speeds and average rates.
Example: A cyclist pedals from his house to his college at speed of 10 km/hr and back from the college to
his house at 15 km/hr. Find the average speed.
Solution: Here the distance is constant
The simple H.M is appropriate for this problem.
X1= 10km/hr X2=15km/hr
2
H.M   12km / hr
1 1

10 15

The Mode
- Mode is a value which occurs most frequently in a set of values
- The mode may not exist and even if it does exist, it may not be unique.
Examples:
6|Page
By Habitamu W. (MSc)
Probability and statistics for Computer Science

1. Find the mode of 5, 3, 5, 8, 9

Mode =5
2. Find the mode of 8, 9, 9, 7, 8, 2, and 5.
It is a bimodal Data: 8 and 9
3. Find the mode of 4, 12, 3, 6, and 7.
No mode for this data.
- The mode of a set of numbers X1, X2, Xn is usually denoted by X̂
Mode for Grouped data
If data are given in the shape of continuous frequency distribution, the mode is defined as:

 1 
X̂  L mo  w 
 1   2 
Where:
Xˆ  the mod e of the distribution
w  the size of the mod al class
 1  f mo  f 1
 2  f mo  f 2
f mo  frequencyof the mod al class
f 1  frequencyof the class preceedingthe mod al class
f 2  frequencyof the class following the mod al class
Note: The modal class is a class with the highest frequency.
Example: Following is the distribution of the size of certain farms selected at random from a
district. Calculate the mode of the distribution.

Size of farms No. of farms

5-15 8
15-25 12
25-35 17
35-45 29
45-55 31
55-65 5
65-75 3
Solutions:

7|Page
By Habitamu W. (MSc)
Probability and statistics for Computer Science

45  55 is the mod al class, sin ce it is a class with the highest frequency.

Lmo  45
w  10
 1  f mo  f 1  2
 2  f mo  f 2  26
f mo  31
f 1  29
f2  5

 Xˆ  45  10
2 

 2  26 
 45.71
Merits and Demerits of Mode
Merits:
 It is not affected by extreme observations.
 Easy to calculate and simple to understand.
 It can be calculated for distribution with open end class
Demerits:
 It is not rigidly defined.
 It is not based on all observations
 It is not suitable for further mathematical treatment.
 It is not stable average
 Often its value is not unique.
The Median
-Median is denoted by X̂ .
Median for ungrouped data

 X ( n1) 2  ,If n is odd.

~ 1
X   (X  X ), If n is even
 2 n 2  ( n 2 )  1
 
 


Example: Find the median of the following numbers.

a) 6, 5, 2, 8, 9, 4.
b) 2, 1, 8, 3, 5, 8.
Solutions:
a) First order the data: 2, 4, 5, 6, 8, 9
Here n=6

8|Page
By Habitamu W. (MSc)
Probability and statistics for Computer Science

~ 1
X  (X n  X n )
2 [2] [  1]
2

1
 ( X [3]  X [ 4 ] )
2
1
 ( 5  6)  5.5
2
b) Order the data :1, 2, 3, 5, 8
Here n=5
~ X
X n 1
[ ]
2

 X[3]
3
Median for grouped data
If data are given in the shape of continuous frequency distribution, the median is defined as:
~ w n
X  L med  (  c)
f med 2
Where :
L med  lower class boundary of the median class.
w  the size of the median class
n  total number of observations.
c  the cumulativefrequency( less than type) preceeding the median class.
f med  thefrequency of the median class.

Remark:
The median class is the class with the smallest cumulative frequency (less than type) greater than or
n
equal to .
2
Example: Find the median of the following distribution.

Class Frequency
40-44 7
45-49 10
50-54 22
55-59 15
60-64 12
65-69 6
70-74 3
Solutions:
 First find the less than cumulative frequency.

9|Page
By Habitamu W. (MSc)
Probability and statistics for Computer Science

 Identify the median class.

 Find median using formula.

Class Frequency Cumu.Freq(less

than type)
40-44 7 7
45-49 10 17
50-54 22 39
55-59 15 54
60-64 12 66
65-69 6 72
70-74 3 75

n 75
  37.5
2 2
39 is the first cumulative frequencyto be greater thanor equalto 37.5
 50  54 is the median class.

L  49.5, w  5
med
n  75, c  17, f  22
med

~
 X L  w ( n  c)
med f 2
med
 49.5  5 (37.5  17)
22
 54.16
Merits and Demerits of Median

Merits:
 Median is a positional average.
 Not influenced by extreme observations.
 Can be calculated in the case of open end intervals.
 Median can be located even if the data are incomplete.
Demerits:
 It is not a good representative of data if the number of items is small.
 It is not applicable for further algebraic treatment.
 It is susceptible to sampling fluctuations.
Quintiles
Quartiles:
- Quartiles are measures that divide the frequency distribution in to four equal parts.

10 | P a g e
By Habitamu W. (MSc)
Probability and statistics for Computer Science

- Q1 is a value which has 25% items which are less than or equal to it. Similarly Q2 has
50%items with value less than or equal to it and Q3 has 75% items whose values are less than
or equal to it.
iN
- To find Qi (i=1, 2, 3) we count of the classes beginning from the lowest class.
4
- For grouped data: we have the following formula
Q L  w ( iN  c) ,i  1,2,3
i Qi f
Q 4
i
Where :
L Q  lower class boundary of thequartile class.
i
w  thesize of thequartile class
N  total number of observatio ns.
c  thecumulativefrequency (less thantype) preceeding thequartile class.
f Q  thefrequency of thequartile class.
i

Remark:
The quartile class (class containing Qi) is the class with the smallest cumulative frequency (less than
iN
type) greater than or equal to .
4
Deciles:
- Deciles are measures that divide the frequency distribution in to ten equal parts.
iN
- To find Di (i=1, 2,..9) we count of the classes beginning from the lowest class.
10
- For grouped data: we have the following formula
w iN
Di  LD i  (  c) , i  1,2,..., 9
f Di 10
Where :
LDi  lower class boundaryof the decile class.
w  the size of the decileclass
N  total number of observations.
c  the cumulative frequency(less thantype) preceedingthe decile class.
f Di  thefrequency of the decile class.

Remark:

11 | P a g e
By Habitamu W. (MSc)
Probability and statistics for Computer Science

The deciles class (class containing Di) is the class with the smallest cumulative frequency (less than
iN
type) greater than or equal to .
10
Percentiles:
- Percentiles are measures that divide the frequency distribution in to hundred equal parts.
- The values of the variables corresponding to these divisions are denoted P1, P2,.. P99 often
called the first, the second,…, the ninety-ninth percentile respectively.
iN
- To find Pi (i=1, 2,..99) we count of the classes beginning from the lowest class.
100
- For grouped data: we have the following formula
w iN
P L  (  c) , i  1,2,..., 99
i P
i f P 100
i
Where :
L  lower class boundary of the percentile class.
P
i
w  the size of the percentile class
N  total number of observations.
c  the cumulative frequency (less than type) preceeding the percentile class.
f  thefrequency of the percentile class.
P
i

Remark:
The percentile class (class containing Pi) is the class with the small cumulative frequency
iN
(less than type) greater than or equal to .
100
Example: Considering the following distribution

Calculate:
a) All quartiles.
b) The 7th decile.
c) The 90th percentile.

Values Frequency
140- 150 17
150- 160 29
160- 170 42
170- 180 72
12 | P a g e
By Habitamu W. (MSc)
Probability and statistics for Computer Science

180- 190 84
190- 200 107
200- 210 49
210- 220 34
220- 230 31
230- 240 16
240- 250 12

Solutions:
 First find the less than cumulative frequency.
 Use the formula to calculate the required quantile.
Values Frequency Cum.Freq(less
than type)
140- 150 17 17
150- 160 29 46
160- 170 42 88
170- 180 72 160
180- 190 84 244
190- 200 107 351
200- 210 49 400
210- 220 34 434
220- 230 31 465
230- 240 16 481
240- 250 12 493

a) Quartiles:
i. Q1
- determine the class containing the first quartile.
N
 123.25
4
 170  180 is the class containingthe first quartile.

w N
 Q1  LQ1  (  c)
LQ  170 , w 10 fQ 4
1
1

N  493 , c  88 , f Q  72 10
1  170  (123.25  88)
72
 174.90

ii. Q2
- determine the class containing the second quartile.

13 | P a g e
By Habitamu W. (MSc)
Probability and statistics for Computer Science

2* N
 246.5
4
 190  200 is the class containing the sec ond quartile.

LQ  190 ,
2
w 10
N  493 , c  244 , f Q 107
2

w 2* N
 Q2  LQ  (  c)
2
fQ 2
4
10
 170  (246.5  244)
72
 190.23

iii. Q3
- determine the class containing the third quartile.
3* N
 369.75
4
 200  210 is the class containing the third quartile.

LQ  200 ,
3
w 10
N  493 , c  351 , f Q  49
3

b) D7
- determine the class containing the 7th decile.
7* N
 345.1
10
190  200 is the class containing the seventh decile.
LD  190 ,
7
w 10
N  493 , c  244 , f D 107
7

14 | P a g e
By Habitamu W. (MSc)
Probability and statistics for Computer Science

w 7* N
 D7  LD  (  c)
7
f D 10
7

10
 190  (345.1  244)
107
 199.45

c) P90
- determine the class containing the 90th percentile.
90 * N
 443.7
100
 220  230 is the class containing the 90th percentile .
LP  220 ,
90
w 10
N  493 , c  434 , f P  3107
90

w 90 * N
 P90  LP  (  c)
90
f P 100
90

10
 220  (443.7  434)
31
 223.13

15 | P a g e
By Habitamu W. (MSc)