MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

CHAPTER - 2
2. Summarization of Data
2.1 Measures of Central Tendency
The most important objective of a statistical analysis is to determine a single value for the
entire mass of data, which describes the overall level of the group of observations and can be
called a representative of the whole set of data. It tells us where the center of the distribution
of data is located. The most commonly used measures of central tendencies are:
 The Mean (Arithmetic mean, Weighted mean, Geometric mean and Harmonic
means)
 The Mode
 The Median

The Summation Notation:

 Let 𝑋1 , 𝑋2 , … , 𝑋𝑁 be the number of measurements where 𝑁 is the total number of

observation and 𝑋𝑖 , is the ith observation.
 Very often in statistics an algebraic expression of the form 𝑋1 + 𝑋2 + ⋯ + 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 ∑𝑁 𝑖=1 𝑋𝑖 is a mathematical shorthand for 𝑋1 + 𝑋2 + ⋯ + 𝑋𝑁 .
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒, ∑𝑵𝒊=𝟏 𝑿𝒊 = 𝑿𝟏 + 𝑿𝟐 + ⋯ + 𝑿𝑵 → 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛𝑠.

∑𝒏𝒊=𝟏 𝑿𝒊 = 𝑿𝟏 + 𝑿𝟐 + ⋯ + 𝑿𝒏 → 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑠𝑎𝑚𝑝𝑙𝑒𝑠.

Example: Suppose the following were scores (marks) made on the first assignment for five
students in the class: 5, 7, 7, 6, 𝑎𝑛𝑑 8. Write their marks using summation notation.
Solution: ∑5𝑖=1 𝑋𝑖 = 𝑋1 + 𝑋2 + ⋯ + 𝑋5 = 5 + 7 + 7 + 6 + 8 = 33
Properties of summation
1. ∑𝑛𝑖=1 𝑐 = 𝑛𝑐 𝑓𝑜𝑟 𝑎𝑛𝑦 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝑐.
𝑛 𝑛
2. ∑𝑖=1 𝑐𝑋𝑖 = 𝑐 ∑𝑖=1 𝑋𝑖
3. ∑𝑛𝑖=1(𝑋𝑖 ± 𝑐) = ∑𝑛𝑖=1 𝑋𝑖 ± 𝑛𝑐
4. ∑𝑛𝑖=1(𝑌𝑖 +𝑋𝑖 ) = ∑𝑛𝑖=1 𝑋𝑖 + ∑𝑛𝑖=1 𝑌𝑖
5. ∑𝑛𝑖=1 𝑋𝑖 𝑌𝑖 = 𝑋1 𝑌1 + 𝑋2 𝑌2 + ⋯ + 𝑋𝑛 𝑌𝑛
𝑛(𝑛+1)
6. 1 + 2+ 3 + 4+ ⋯+ 𝑛 = 2
𝑛(𝑛+1)(2𝑛+1)
7. 12 + 22 +32 + ⋯ + 𝑛2 = 6
2 2
 N. B ∑Xi ≠ (∑Xi ) and (∑XiYi) ≠ (∑Xi)(∑Yi)
2.2 Types of Measure of Central Tendency
2.2.1 The Mean
2.2.1.1 Arithmetic mean
The arithmetic mean of a sample is the sum of all observations divided by the number of
observations in the sample. i.e.
𝒕𝒉𝒆 𝒔𝒖𝒎 𝒐𝒇 𝒂𝒍𝒍 𝒗𝒂𝒍𝒖𝒆𝒔 𝒊𝒏 𝒕𝒉𝒆 𝒔𝒂𝒎𝒑𝒍𝒆
𝑺𝒂𝒎𝒑𝒍𝒆 𝒎𝒆𝒂𝒏 𝒐𝒓 𝒂𝒓𝒊𝒕𝒉𝒎𝒆𝒕𝒊𝒄 𝒎𝒆𝒂𝒏 = 𝒏𝒖𝒎𝒃𝒆𝒓 𝒐𝒇 𝒗𝒂𝒍𝒖𝒆𝒔 𝒊𝒏 𝒕𝒉𝒆 𝒔𝒂𝒎𝒑𝒍𝒆

By Admasu M. Lecture Notes Page 1

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

Suppose that 𝑥1 , 𝑥2 , … , 𝑥𝑛 are n observed values in a sample of size n taken from a

population of size N. Then the arithmetic mean of the sample, denoted by 𝑥̅ , is given by
𝑿𝟏 +𝑿𝟐 +⋯+𝑿𝒏 ∑𝒏
𝒊=𝟏 𝑿𝒊
̅=
𝑿 = → 𝐟𝐨𝐫 𝐬𝐚𝐦𝐩𝐥𝐞𝐬.
𝐧 𝒏
If we take an entire population, the population mean denoted by µ is given by
𝑿𝟏 +𝑿𝟐 +⋯+𝑿𝑵 ∑𝑵
𝒊=𝟏 𝑿𝒊
µ= = → 𝐟𝐨𝐫 𝐩𝐨𝐩𝐮𝐥𝐚𝐭𝐢𝐨𝐧𝐬.
𝐍 𝑵
In general, the sample arithmetic mean is calculated by

𝑿𝒊
∑𝒏𝒊=𝟏 → 𝑓𝑜𝑟 𝑟𝑎𝑤 𝑑𝑎𝑡𝑎
𝒏
𝒌 𝒇𝒊 𝑿𝒊
̅=
𝑿 ∑𝒊=𝟏 → 𝑓𝑜𝑟 𝑢𝑛𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎 𝑤ℎ𝑒𝑟𝑒 ∑ 𝑓𝑖 = 𝑛.
∑ 𝒇𝒊
𝒌 𝒇𝒊 𝑴 𝒊
{∑𝒊=𝟏 ∑ 𝒇𝒊 → 𝑓𝑜𝑟 𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎 𝑤ℎ𝑒𝑟𝑒 ∑ 𝑓𝑖 = 𝑛.
Example 1: The net weights of five perfume bottles selected at random from the production
line 𝑎𝑟𝑒 85.4, 85.3, 84.9, 85.4 𝑎𝑛𝑑 85.what is the arithmetic mean weight of the sample
observation?
Solution; 𝐺𝑖𝑣𝑒𝑛 𝑛 = 5 𝑥1 = 85.4, 𝑥2 = 85.3, 𝑥3 = 84.9, 𝑥4 = 85.4 𝑎𝑛𝑑 𝑥5 =
85.
∑ 𝑛
𝑋𝑖 85.4+85.3+84.9+85.4+ 85 426.6
𝑋̅ = 𝑖=1 = = = 85.32.
𝑛 5 5
Example 2: Calculate the mean of the marks of 46 students given below;

Marks (𝑋𝑖 ) 9 10 11 12 13 14 15 16 17 18
Frequency (𝑓𝑖 ) 1 2 3 6 10 11 7 3 2 1

Solution: ∑ 𝑓𝑖 = 𝑛 = 46 is the sum of the frequencies or total number of observations.

To calculate ∑𝑘𝑖=1 𝒇𝒊 𝑿𝒊 consider the following table.

𝑋𝑖 9 10 11 12 13 14 15 16 17 18 Total
𝑓𝑖 1 2 3 6 10 11 7 3 2 1 46
𝒇𝒊 𝑿𝒊 9 20 33 72 130 154 105 48 34 18 623

𝒇𝑿 623
So 𝑋̅ = ∑𝒌𝒊=𝟏 ∑𝒊 𝒇 𝒊 = 46 = 13.54.
𝒊
Example 3: The net income of a sample of large importers of Urea was organized into the
following table. What is the arithmetic mean of net income?

Net income 2-4 5-7 8-10 11-13 14-16

Number of importers 1 4 10 3 2

Solution: ∑ 𝑓𝑖 = 𝑛 = 20 is the sum of the frequencies or total number of observations.

To calculate ∑𝑘𝑖=1 𝒇𝒊 𝒎𝒊

consider the following table.

By Admasu M. Lecture Notes Page 2

 MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

Net income (CI) 2-4 5-7 8-10 11-13 14-16 Total

Number of importers (𝑓𝑖 ) 1 4 10 3 2 20
Class marks (𝑚𝑖 ) 3 6 9 12 15
𝒇𝒊 𝒎𝒊 3 24 90 36 30 183

𝒇𝒊 𝒎 𝒊 𝟏𝟖𝟑
So 𝑋̅ = ∑𝒌𝒊=𝟏 ∑ 𝒇𝒊
= = 𝟗. 𝟏𝟓.
𝟐𝟎
Example 4: From the following data, calculate the missing frequency? The mean number
of tablets to cure ever was 29.18.

Number of tablets 19 − 21 22 − 24 25 − 27 28 − 30 31 − 33 34 − 36 37 − 39
Number of 6 13 19 𝑓4 18 12 9
persons cured

Solution; ∑ 𝑓𝑖 = 𝑛 = 77 + 𝑓4 is the sum of the frequencies or total number of

observations.
To calculate ∑𝑘𝑖=1 𝒇𝒊 𝒎𝒊 consider the following table.

CI 19 − 21 22 − 24 25 − 27 28 − 30 31 − 33 34 37 − 39 Total
− 36
𝒇𝒊 6 13 19 𝑓4 18 12 9 77+𝑓4
𝒎𝒊 20 23 26 29 32 35 38
𝒇𝒊 𝒎𝒊 120 299 494 29𝑓4 576 420 342 2251
+ 29𝑓4

𝑆𝑖𝑛𝑐𝑒 𝑚𝑒𝑎𝑛 𝑖𝑠 𝑔𝑖𝑣𝑒𝑛 𝑡𝑜 𝑏𝑒 29.18 𝑤𝑒 ℎ𝑎𝑣𝑒

𝑘
𝑓𝑖 𝑚𝑖 2251 + 29𝑓4
𝑋̅ = ∑ =≫ 29.18 = =≫ 29.18 (77+𝑓4 )
∑ 𝑓𝑖 77+𝑓4
𝑖=1
= 2251 + 29𝑓4
4.14
=≫ 29.18𝑓4 − 29𝑓4 = 2251 − 2246.86 =≫ 0.18𝑓4 = 4.14 =≫ 𝑓4 = 0.18 = 23.
Combined mean
If we have an arithmetic means 𝑋1 , 𝑋2 , … , 𝑋𝑛 of n groups having the same unit of
measurement of a variable, with sizes 𝑛1 , 𝑛2 , … , 𝑛𝑛 observations respectively, we can
compute the combined mean of the variant values of the groups taken together from the
individual means by
̅𝟏 +𝒏𝟐 𝒙
𝒏𝟏 𝒙 ̅𝟐 +⋯+𝒏𝒊 𝒙
̅𝒏 ∑𝒏 ̅𝒊
𝒊=𝟏 𝒏𝒊 𝒙
𝑿𝒄𝒐𝒎 = = ∑𝒏
𝒏𝟏 +𝒏𝟐 +⋯+𝒏𝒊 𝒊=𝟏 𝒏𝒊
Example 1: Compute the combined mean for the following two sets.
𝑺𝒆𝒕 𝑨: 1, 4, 12, 2, 8 𝑎𝑛𝑑 6 ; 𝑺𝒆𝒕 𝑩: 3, 6, 2, 7 𝑎𝑛𝑑 4.
∑6𝑖=1 𝑋𝑖 33 ∑5𝑖=1 𝑋𝑖 22
Solution: 𝑛1 = 6, 𝑥̅1 = = = 5.5 ; 𝑛2 = 6, 𝑥̅2 = = =
𝒏𝟏 6 𝒏𝟐 5
4.4.

By Admasu M. Lecture Notes Page 3

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

𝑛1 𝑥̅1 + 𝑛2 𝑥̅2 6 𝑥 5.5 + 5 𝑥 4.4 55

𝑋𝑐𝑜𝑚 = = = = 5.
𝑛1 + 𝑛2 6+5 11
Example 2: The mean weight of 150 students in a certain class is 60 kg. The mean
weight of boys in the class is 70 kg and that of girl’s is 55 kg . Find the number of
boys and girls in the class?
Solution; Let 𝑛1 be the number of boys and 𝑛2 be the number of girls in the class.
Also let 𝑥1 , 𝑥2 𝑎𝑛𝑑 𝑥𝑐𝑜𝑚 be the mean weights of boys, girls and the mean weights of all
students respectively. Then 𝑥1 = 70, 𝑥2 = 55 𝑎𝑛𝑑 𝑥𝑐𝑜𝑚 = 60. 𝐶𝑙𝑒𝑎𝑟𝑙𝑦 𝑛1 + 𝑛2 =
𝑛1 𝑥̅ 1 +𝑛2 𝑥̅ 2 70 𝑛1 +55𝑛2 70 𝑛1 +55𝑛2
150. 𝑋𝑐𝑜𝑚 = =≫ 60 = =≫ 60 =
𝑛1 +𝑛2 𝑛1 +𝑛2 150
=≫ 9000 = 70 𝑛1 + 55𝑛2 … . … (1) 𝑎𝑛𝑑
150 = 𝑛1 + 𝑛2 …… (2)
𝑆𝑜𝑙𝑣𝑖𝑛𝑔 𝑒𝑞𝑢𝑎𝑡𝑖𝑜𝑛 (1) 𝑎𝑛𝑑 (2) 𝑠𝑖𝑚𝑢𝑙𝑡𝑎𝑛𝑒𝑜𝑢𝑠𝑙𝑦, 𝑤𝑒 𝑔𝑒𝑡 𝑛1 = 50 𝑎𝑛𝑑 𝑛2 = 100.
Disadvantages of the arithmetic mean
1. The mean is meaningless in the case of nominal or qualitative data.
2. In case of grouped data, if any class interval is open ended, arithmetic mean cannot be
calculated, since the class mark of this interval cannot be found.
2.2.1.2 Weighted mean
In the computation of arithmetic mean, we had given an equal importance to each
observation. Sometimes the individual values in the data may not have an equally
importance. When this is the case, we assigned to each weight which is proportional to its
relative importance.
 The weighted mean of a set of values 𝑥1 , 𝑥2 , … , 𝑥𝑛 with corresponding weights
𝑤1 , 𝑤2 , … , 𝑤𝑛 denoted by 𝑥̅𝑤 is computed by:
𝒘𝟏 𝒙𝟏 + 𝒘𝟐 𝒙𝟐 + ⋯ + 𝒘𝒏 𝒙𝒏 ∑𝒏𝒊=𝟏 𝒘𝒊 𝒙𝒊
𝑿𝒘 = = 𝒏
𝒘𝟏 + 𝒘𝟐 + ⋯ + 𝒘𝒏 ∑𝒊=𝟏 𝒘𝒊
The calculation of cumulative grade point average (CGPA) in Colleges and Universities is a
good example of weighted mean.
Example: If a student scores "A" in a 3 EtCTS course, "B" in a 6 EtCTS course, "B" in
another 5 EtCTS course and "D" in a 2 EtCTS course. Compute his /her GPA for the
semester.
Solution: Here the numerical values of the letter grades are the values (i.e. 𝐴 =
4, 𝐵 = 3, 𝐶 = 2 𝑎𝑛𝑑 𝐷 = 1) and the corresponding EtCTS of the course are their
respective weights. i.e.

Grade values (𝒙𝒊 ) 4 3 3 1

Weight (𝒘𝒊 ) 3 6 5 2

𝒘𝟏 𝒙𝟏 +𝒘𝟐 𝒙𝟐 +⋯+𝒘𝒏 𝒙𝒏 ∑𝟒𝒊=𝟏 𝒘𝒊 𝒙𝒊 4x3+3x6+3x5+1x2 12+18+15+2 47

𝑮𝑷𝑨 = 𝑿𝒘 = = ∑𝟒𝒊=𝟏 𝒘𝒊
= = = 16 =
𝒘𝟏 +𝒘𝟐 +⋯+𝒘𝒏 3+6+5+2 16
2.9375.
2.2.1.3 Geometric mean
In algebra geometric mean is calculated in the case of geometric progression, but in statistics
we need not bother about the progression, here it is particular type of data for which the

By Admasu M. Lecture Notes Page 4

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

geometric mean is of great importance because it gives a good mean value. If the observed
values are measured as ratios, proportions or percentages, then the geometric mean gives a
better measure of central tendency than any other means.
 The Geometrical mean of a set of values 𝑥1 , 𝑥2 , … , 𝑥𝑛 of n positive values is defined as
the nth root of their product . That is,
𝐺. 𝑀 = 𝑛√ 𝑥1 ∗ 𝑥2 ∗ … ∗ 𝑥𝑛
Example: The G.M of 4, 8 and 6 is
3 3
𝐺. 𝑀 = √ 4 𝑥 8 𝑥 6 = √ 192 = 5.77.
In general, the sample geometric mean is calculated by

𝑮. 𝒎 =
𝑛
√ 𝑥1 ∗ 𝑥2 ∗ … ∗ 𝑥𝑛 → 𝑓𝑜𝑟 𝑟𝑎𝑤 𝑑𝑡𝑎𝑡𝑎
{ 𝑛√ 𝑥1 𝑓1 ∗ 𝑥2 𝑓2 ∗, … ,∗ 𝑥𝑘 𝑓𝑘 → 𝑓𝑜𝑟 𝑢𝑛𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎 𝑤ℎ𝑒𝑟𝑒 𝑛 = ∑ 𝑓𝑖 .
𝑛
√ 𝑚1 𝑓1 ∗ 𝑚2 𝑓2 ∗. , … ,∗ 𝑚𝑘 𝑓𝑘 → 𝑓𝑜𝑟 𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎 𝑤ℎ𝑒𝑟𝑒 𝑛 = ∑ 𝑓𝑖 .
Example1: The man gets three annual raises in his salary. At the end of first year, he
gets an increase of 4%, at the end of the second year, he gets an increase of 6% and at
the
end of the third year, he gets an increase of 9% of his salary. What is the average
percentage increase in the three periods?
3
Solution: 𝐺. 𝑀 = √1.04 ∗ 1.06 ∗ 1.09 = 1.0631 => 1.0631 − 1 =
0.0631.
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒, 𝑡ℎ𝑒 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝑖𝑛𝑐𝑟𝑒𝑎𝑠𝑒 𝑖𝑠 6.31%.
Example 2: Compute the Geometric mean of the following data.

Values 2 4 6 8 10
Frequency 1 2 2 2 1
8 8 8
𝑺𝒐𝒍𝒖𝒕𝒊𝒐𝒏: 𝑔. 𝑚 = √ 21 ∗ 42 ∗ 62 ∗ 82 ∗ 101 = √2 ∗ 16 ∗ 36 ∗ 64 ∗ 100 = √737280 =
5.41. Example 3: Suppose that the profits earned by a certain construction company in four
projects were 3%, 2%, 4% & 6% respectively. What is the Geometric mean profit?
4 4
𝑺𝒐𝒍𝒖𝒕𝒊𝒐𝒏: 𝑔. 𝑚 = √3 ∗ 2 ∗ 4 ∗ 6 = √144 = 3.46.
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒; 𝑡ℎ𝑒 𝑔𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐 𝑚𝑒𝑎𝑛 𝑝𝑟𝑜𝑓𝑖𝑡 𝑖𝑠 3.46 𝑝𝑒𝑟𝑐𝑒𝑛𝑡.
2.2.1.4 Harmonic mean
Another important mean is the harmonic mean, which is suitable measure of central tendency
when the data pertains to speed, rates and price.
 Let 𝑥1 , 𝑥2 , … , 𝑥𝑛 be n variant values in a set of observations, then simple harmonic
𝐧 𝐧
mean is given by: 𝑺. 𝑯. 𝑴 = 𝟏 𝟏 𝟏 = 𝐧 𝟏
+ +⋯+ ∑𝐢=𝟏
𝐱𝟏 𝐱𝟐 𝐱𝐧 𝐱𝐢

 Note: SHM is used for equal distances, equal costs and equal rates.

Example 1: A motorist travels for three days at a rate (speed) of 480 km/day. On the first
day he travels 10 hours at a rate of 48 km/h, on the second day 12 hours at a rate of 40 km/h,
on the third day 15 hours at a rate of 32 km/h. What is the average speed?

By Admasu M. Lecture Notes Page 5

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

Solution: Since the distance covered by the motorist is equal (𝑖. 𝑒. 𝑠1 = 480, 𝑠2 = 480, 𝑠3 =
480), so we use SHM.
3
𝑆. 𝐻. 𝑀 = 1/48+1/40+1/32 = 38.92 so the required average speed = 38.92 𝑘𝑚/ℎ.
We can check this, by using the known formula for average speed in elementary physics.
total distance covered 𝑆𝑇
Check; 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑠𝑝𝑒𝑒𝑑 (𝑉𝑎𝑣 ) = =
total time taken 𝑡𝑇
480km+480km+480km 1440km
= = = 38.92 𝑘𝑚/ h.
10hr+12hr+15hr 37hr
Example 2: A business man spent 20 Birr for milk at 40 cents per liter in Mizan-Aman town
and another 20 Birr at 60 cents per liter in Tepi town. What is the average price of milk at
two towns.
Solution: Since the price on the two towns are equal (20 Birr), so we use
SHM.
2
𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑝𝑟𝑖𝑐𝑒 (𝑝𝑎𝑣 ) = 𝑆. 𝐻. 𝑀 = 1 1 = 48 𝑐𝑒𝑛𝑡𝑠/𝑙𝑖𝑡𝑒𝑟.
+
40 60

Weighted harmonic mean (WHM)

 WHM is used for different distance, different cost and different rate.
∑ 𝑤𝑖
𝑊. 𝐻. 𝑀 = w
∑ i
xi
st
Example 1: A driver travel for 3 days. On the 1 day he drives for 10h at a speed of 48 km/h,
on the 2nd day for 12h at 45 km/h and on the 3rd day for 15h at 40 km/h. What is the average
speed?
Solution: since the distance covered by the driver is not equal, so we use WHM by
taking the distance as weights (wi).
∑ 𝑤𝑖 (480 + 540 + 600)𝑘𝑚
𝑣𝑎𝑣 = 𝑤. ℎ. 𝑚 = w = = 43.32 𝑘𝑚/ℎ𝑟.
∑ i (480 + 540 + 600) ℎ𝑟
xi 40 45 40
Example 2: A business man spent 20 Birr for milk at rate of 40 cents per liter in Mizan-
Aman town and 25 Birr at a price of 50 cents per liter in Tepi town. What is the average price
?

Solution: Since the price on the two towns are different , so we use WHM by taking the cost
as weights (wi).
∑ 𝑤𝑖 (20 + 25) 𝑏𝑖𝑟𝑟
𝑝𝑎𝑣 = 𝑤. ℎ. 𝑚 = w = = 45 𝑐/𝑙.
∑ i (20 + 25) 𝑏𝑖𝑟𝑟 𝑙/𝑐
xi 40 50
 (Finally if all the observations are positive) 𝐴. 𝑀 ≥ 𝐺. 𝑀 ≥ 𝐻. 𝑀.
Corrected mean
𝒄−𝒘
̅𝒄𝒐𝒓𝒓 = 𝒙
𝒙 ̅𝒘 + 𝑤ℎ𝑒𝑟𝑒 𝑥̅𝑤 𝑖𝑠 𝑡ℎ𝑒 𝑤𝑟𝑜𝑛𝑔 𝑚𝑒𝑎𝑛
𝒏

By Admasu M. Lecture Notes Page 6

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

 𝒄 𝑖𝑠 𝑡ℎ𝑒 𝑠𝑢𝑚 𝑜𝑓 𝑐𝑜𝑟𝑟𝑒𝑐𝑡 𝑣𝑎𝑙𝑢𝑒𝑠 𝑎𝑛𝑑 𝒘 𝑖𝑠 𝑡ℎ𝑒 𝑠𝑢𝑚 𝑜𝑓 𝑤𝑟𝑜𝑛𝑔 𝑣𝑎𝑙𝑢𝑒𝑠.

Example: The mean age of a group of 100 persons was found to be 32.02 years. Later on, it
was discovered that age of 57 was misread as 27. Find the corrected mean?
Solution: 𝑛 = 100, 𝑥̅𝑤 = 32.02 , 𝑐 = 57 𝑎𝑛𝑑 𝑤 = 27.
𝑐−𝑤 57 − 27
𝑥̅ 𝑐𝑜𝑟𝑟 = 𝑥̅𝑤 + = 32.02 + = 32.02 + 0.3 = 32.32 𝑦𝑒𝑎𝑟𝑠.
𝑛 100
Median and mode
2.2.2 The Median
Suppose we sort all the observations in numerical order, ranging from smallest to largest or
vice versa. Then the median is the middle value in the sorted list. We denote it by x̃.
Let 𝑥1 , 𝑥2 , … , 𝑥𝑛 be n ordered observations. Then the median is given by:
𝑿 𝒏+𝟏 𝐼𝑓 𝑛 𝑖𝑠 𝑜𝑑𝑑.
𝟐
̃ = { 𝑿 𝒏 +𝑿 𝒏+𝟏
𝒙
𝟐 𝟐
𝐼𝑓 𝑛 𝑖𝑠 𝑒𝑣𝑒𝑛.
𝟐
Example 1: Find the median for the following data.
23, 16, 31, 77, 21, 14, 32, 6, 155, 9, 36, 24, 5, 27, 19
Solution: First arrange the given data in increasing order. That is
5, 6, 9, 14, 16, 19, 21, 23, 24, 27, 31, 32, 36, 77, 𝑎𝑛𝑑 155.
𝑛 = 15 =≫ 𝑜𝑑𝑑, 𝑥̃ = 𝑋 𝑛+1 = 𝑋 15+1 = 𝑋 8 = 23
2 2
Example 2: Find the median for the following data.
61, 62, 63, 64, 64, 60, 65, 61, 63, 64, 65, 66, 64, 63
Solution: First arrange the given data in increasing order. that is
60, 61, 61, 62, 63, 63, 63, 64, 64, 64, 65, 65 & 66.
𝑋 𝑛 + 𝑋 𝑛+1 𝑋 14 + 𝑋 14+1 𝑋 + 𝑋 63 + 64 127
7 8
𝑛 = 14 =≫ 𝑒𝑣𝑒𝑛, 𝑥̃ = 2 2
= 2 2
= = =
2 2 2 2 2
= 63.5

Median for ungrouped data

 Let 𝑥1 , 𝑥2 , … , 𝑥𝑘 have their corresponding frequencies 𝑓1 , 𝑓2 , … , 𝑓𝑘 then to find the
median:

 First sort the data in ascending order.

 Construct the less than cumulative frequency (lcf) .
n+1 n+1
 If 𝑛 = ∑𝑘𝑖=1 fi is odd, find and search the smallest lcf which is≥ . Then
2 2
the variant value corresponding to this lcf is the median.
n 𝑛
 If n is even, find 2 & 2 + 1 and search the smallest lcf which is ≥
n 𝑛
(2) & (2 + 1). Then the average of the variant values corresponding to these lcf is
the median.
Example 1: Find the median for the following data.

Values (xi) 3 5 4 2 7 6
Frequency (fi) 2 1 3 2 1 1

By Admasu M. Lecture Notes Page 7

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

Solution: First arrange the data in increasing order and construct the lcf table for this data.

Values (xi) 2 3 4 5 6 7
Frequency (fi) 2 2 3 1 1 1
Lcf 2 4 7 8 9 10

𝑛 10 𝑛 10
𝑛 = 10 =≫ 𝑒𝑣𝑒𝑛. 𝑆𝑜 = = 5 𝑎𝑛𝑑 +1= + 1 = 5 + 1 = 6.
2 2 2 2
Then the smallest LCF which is ≥ 5 & 6 𝑖𝑠 7 and the variant value corresponding to this
4+4
LCF is 4. Thus the median is x̃ = = 4.
2
Example 2: Calculate the median of the marks of 46 students given below.

Values (xi) 10 9 11 12 14 13 15 16 17 18
Frequency (fi) 2 1 3 6 10 11 7 3 2 1

Solution: First arrange the data in ascending order and construct the LCF table for this data.

Values (xi) 9 10 11 12 13 14 15 16 17 18
Frequency (fi) 1 2 3 6 11 10 7 3 2 1
LCF 1 3 6 12 23 33 40 43 45 46

𝑛 46 𝑛 46
𝑛 = 46 =≫ 𝑒𝑣𝑒𝑛. 𝑆𝑜 = = 23 𝑎𝑛𝑑 + 1 = + 1 = 23 + 1 = 24.
2 2 2 2
𝑇ℎ𝑒 𝑠𝑚𝑎𝑙𝑙𝑒𝑠𝑡 𝐿𝐶𝐹 ≥ 23 & 24 𝑎𝑟𝑒 23 & 33 𝑟𝑒𝑠𝑝𝑒𝑐𝑡𝑖𝑣𝑒𝑙𝑦 𝑎𝑛𝑑 the variant values
13+14
corresponding to these LCF are 13 & 14 respectively. Thus the median 𝑖𝑠 x̃ = =
2
13.5.
Median for grouped data
The formula for computing the median for grouped data is given by
𝒏
(𝟐 − 𝒍𝒄𝒇𝒑 ) 𝒙 𝒘
𝒎𝒆𝒅𝒊𝒂𝒏 = 𝐱̃ = 𝐥𝐜𝐛𝒙̃ +
𝒇𝒎
𝑊ℎ𝑒𝑟𝑒: 𝑙𝑐𝑏𝑥̃ − 𝑖𝑠 𝑡ℎ𝑒 𝒍𝒄𝒃 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑒𝑑𝑖𝑎𝑛 𝑐𝑙𝑎𝑠𝑠.
 𝑛 − 𝑖𝑠 𝑡ℎ𝑒 𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑜𝑏𝑠𝑒𝑟𝑣𝑎𝑡𝑖𝑜𝑛𝑠.
 𝑓𝑚 𝑖𝑠 𝑡ℎ𝑒 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑒𝑑𝑖𝑎𝑛 𝑐𝑙𝑎𝑠𝑠.
 𝑤 𝑖𝑠 𝑡ℎ𝑒 𝑤𝑖𝑑𝑡ℎ 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑒𝑑𝑖𝑎𝑛 𝑐𝑙𝑎𝑠𝑠. 𝑟𝑒𝑐𝑎𝑙𝑙 ∶ 𝑤 = 𝑢𝑐𝑏 − 𝑙𝑐𝑏.
 𝑙𝑐𝑓𝑝 𝑖𝑠 𝑡ℎ𝑒 𝒍𝒄𝒇 𝑐𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑡𝑜 𝑡ℎ𝑒 𝑐𝑙𝑎𝑠𝑠 𝒊𝒎𝒎𝒆𝒅𝒊𝒂𝒕𝒆𝒍𝒚 𝒑𝒓𝒆𝒄𝒆𝒅𝒊𝒏𝒈 𝑡ℎ𝑒 𝑚𝑒𝑑𝑖𝑎𝑛 𝑐𝑙𝑎𝑠𝑠.

n
 Note: The class corresponding to the smallest LCF which is ≥ is called the
2
median class. So that the median lies in this class.
Steps to calculate the median for grouped data
1. First construct the LCF table.

By Admasu M. Lecture Notes Page 8

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

n
2. Determine the median class. To determine the median class, find and search the
2
n
smallest LCF which is ≥ 2. Then the class corresponding to this lcf is the median
class.
Example 1: Find the median for the following data.

Daily production 80 − 89 90 − 99 100 − 109 110 − 119 120 − 129 130 − 139
Frequency 5 9 20 8 6 2

Solution: First construct the LCF table.

Daily 80 90 100 110 120 130

production(CI) − 89 − 99 − 109 − 119 − 129 − 139
Frequency(fi) 5 9 20 8 6 2
Lcf 5 14 34 42 48 50

𝑛 50 𝑛
To obtain the median class , calculate = = 25. Thus the smallest lcf which is ≥ is
2 2 2
34. So the class corresponding to this lcf is 100 − 109, 𝑖𝑠 𝑡ℎ𝑒 𝑚𝑒𝑑𝑖𝑎𝑛 𝑐𝑙𝑎𝑠𝑠.
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒, 𝑙𝑐𝑏𝑥̃ = 99.5, 𝑤 = 10, 𝑓𝑚 = 20, 𝑙𝑐𝑓𝑝 = 14.
𝑛
(2 − 𝑙𝑐𝑓𝑝 ) 𝑥 𝑤 (25 − 14) 𝑥 10
𝑚𝑒𝑑𝑖𝑎𝑛 = x̃ = lcb𝑥̃ + = 99.5 + = 105.
𝑓𝑚 20
Properties of the median
1. The median is unique.
2. It can be computed for an open ended frequency distribution if the median does not
lie in an open ended class.
3. It is not affected by extremely large or small values .
4. It is not so suitable for algebraic manipulations.
5. It can be computed for ratio level, interval level and ordinal level data.
2.2.3 The mode
In every day speech, something is “in the mode” if it is fashionable or popular. In statistics
this “popularity” refers to frequency of observations.
Therefore, mode is the `most frequently observed value in a set of observations.
𝑬𝒙𝒂𝒎𝒑𝒍𝒆: 𝑺𝒆𝒕 𝑨: 10, 10, 9, 8, 5, 4, 5, 12, 10 𝑚𝑜𝑑𝑒 = 10 → 𝑢𝑛𝑖𝑚𝑜𝑑𝑎𝑙.
𝑺𝒆𝒕 𝑩: 10, 10, 9, 9, 8, 12, 15, 5 𝑚𝑜𝑑𝑒 = 9 &10 →
𝑏𝑖𝑚𝑜𝑑𝑎𝑙.
𝑺𝒆𝒕 𝑪: 4, 6, 7, 15, 12, 9 𝑛𝑜 𝑚𝑜𝑑𝑒.
Remark: In a set of observed values, all values occur once or equal number of times, there is
no mode. (See set C above).
Mode for a grouped data
If the data is grouped such that we are given frequency distribution of finite class intervals,
we do not know the value of every item, but we easily determine the class with highest
frequency. Therefore, the modal class is the class with the highest frequency. So that the
mode of the distribution lies in this class.

By Admasu M. Lecture Notes Page 9

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

 To compute the mode for a grouped data we use the formula:

∆𝟏
𝒎𝒐𝒅𝒆 = 𝑿 ̂ = 𝒍𝒄𝒃𝒙̂ + ( )𝒙 𝒘
∆𝟏 + ∆𝟐
𝑤ℎ𝑒𝑟𝑒; ∆1 = 𝑓𝑚 − 𝑓𝑝 , ∆2 = 𝑓𝑚 − 𝑓𝑠
𝑊ℎ𝑒𝑟𝑒: 𝑙𝑐𝑏𝑥̂ – 𝑖𝑠 𝑡ℎ𝑒 𝒍𝒄𝒃 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑜𝑑𝑎𝑙 𝑐𝑙𝑎𝑠𝑠.
𝑓𝑚 − 𝑖𝑠 𝑡ℎ𝑒 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑜𝑑𝑎𝑙 𝑐𝑙𝑎𝑠𝑠.

𝑓𝑝 − 𝑖𝑠 𝑡ℎ𝑒 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑐𝑙𝑎𝑠𝑠 𝒑𝒓𝒆𝒄𝒆𝒅𝒊𝒏𝒈 𝑡ℎ𝑒 𝑚𝑜𝑑𝑎𝑙 𝑐𝑙𝑎𝑠𝑠.

𝑓𝑠 − 𝑖𝑠 𝑡ℎ𝑒 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑐𝑙𝑎𝑠𝑠 𝒔𝒖𝒄𝒄𝒆𝒆𝒅𝒊𝒏𝒈 𝑡ℎ𝑒 𝑚𝑜𝑑𝑎𝑙 𝑐𝑙𝑎𝑠𝑠.

𝑤 − 𝑖𝑠 𝑡ℎ𝑒 𝑤𝑖𝑑𝑡ℎ 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑜𝑑𝑎𝑙 𝑐𝑙𝑎𝑠𝑠.
Example 1: The ages of newly hired, unskilled employees are grouped into the following
distribution. Then compute the modal age?

Ages 18 − 20 21 − 23 24 − 26 27 − 29 30 − 32
Number 4 8 11 20 7

Solution: First we determine the modal class. The modal class is 27 − 29, since it has the
highest frequency. 𝑇ℎ𝑢𝑠, 𝑙𝑐𝑏𝑥̂ = 26.5, 𝑤 = 3, ∆1 = 20 − 11 = 9, ∆2 = 20 − 7 = 13.
∆1 9 27
𝑋̂ = 𝑙𝑐𝑏𝑥̂ + ( ) 𝑥 𝑤 = 26.5 + ( ) 𝑥 3 = 26.5 + ( ) = 26.5 + 1.2 = 𝟐𝟕. 𝟕
∆1 + ∆2 9 + 13 22
Interpretation: The age of most of these newly hired employees is 27.7 (27 years and 7
months).
Example 2: The following table shows the distribution of a group of families according to
their expenditure per week. The median and the mode of the following distribution are
known to be 25.50 Birr and 24.50 Birr respectively. Two frequency values are however
missing from the table. Find the missing frequencies.

Class interval 1 − 10 11 − 20 21 − 30 31 − 40 41 − 50
Frequency 14 𝑓2 27 𝑓4 15

Solution: The LCF table of the given distribution can be formed as follows.

Expenditure (CI) 1 − 10 11 − 20 21 − 30 31 − 40 41 − 50
Number of families (fi) 14 𝑓2 27 𝑓4 15
LCF 14 14 + 𝑓2 41 + 𝑓2 41 + 𝑓2 + 𝑓4 56 + 𝑓2 + 𝑓4

Here: 𝑛 = 56 + 𝑓2 + 𝑓4 . Since the median and the mode are Birr 25.5 & 24.5 respectively
then the class 21 − 30 is the median class as well as the modal class.
56+𝑓2 +𝑓4
( −(14+𝑓2 ) ) x 10
2
25.5 = 20.5 + (𝑖)
27
(27−𝑓2 ) x 10
24.5 = 20.5 + (27−𝑓 )+(27−𝑓 ) (𝑖𝑖)
2 4

By Admasu M. Lecture Notes Page 10

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

𝐸𝑞𝑛. (𝑖) 𝑎𝑛𝑑 𝑒𝑞𝑛. (𝑖𝑖) 𝑐𝑎𝑛 𝑏𝑒 𝑤𝑟𝑖𝑡𝑡𝑒𝑛 𝑎𝑠

5 x (56+𝑓2 +𝑓4 )−10 x (14+𝑓2 ) (27−𝑓2 ) x 10
5= & 4=
27 54−𝑓2 −𝑓4
Further simplifying the above we get
𝑓2 − 𝑓4 = 1. (𝑖𝑖𝑖) &
3𝑓2 − 2𝑓4 = 27. (𝑖𝑣)
𝑆𝑜𝑙𝑣𝑖𝑛𝑔 (𝑖𝑖𝑖) & (𝐼𝑉), 𝑤𝑒 𝑔𝑒𝑡 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒𝑠 𝑜𝑓 𝑓2 & 𝑓4 𝑎𝑠 𝑓2 = 25 & 𝑓4 = 24.
Properties of mode
1. it is not affected by extreme values.
2. It can be calculated for distribution with open ended classes.
3. It can be computed for all levels of data i.e. nominal, ordinal, interval and ratio.
4. The main drawback of mode is that often it does not exist.
5. Often its values are not unique.
2.3 Measure of non - central location (Quintiles’)
There are three types of quintiles. These are:
1. Quartiles
The quartiles are the three points, which divide a given order data into four equal parts. These
𝑖 𝑥 (𝑛+1)𝑡ℎ
𝑄𝑖 = , 𝑖 = 1, 2, 3. → 𝑓𝑜𝑟 𝑟𝑎𝑤 𝑑𝑎𝑡𝑎.
4
n+1 th
Q1 is the value corresponding to ( ) ordered observation.
4
n+1 th
Q2 is the value corresponding to 2 𝑥 ( ) ordered observation.
4
n+1 th
Q3 is the value corresponding to 3 𝑥 ( 4 ) ordered observation.
Example: Consider the age data given below and calculate Q1, Q2, and Q3.
19, 20, 22, 22, 17, 22, 20, 23, 17, 18
Solution: First arrange the data in ascending order, n=10.
17, 17, 18, 19, 20, 20, 22, 22, 22, 23
n+1 th 10+1 th
Q1 = ( ) =( ) = (2.75)th observation = 2nd observation + 0.75 (3rd - 2nd)
4 4
observation = 17 + 0.75(18 − 17) = 17.75
Therefore 25% of the observations are below 17.75
n+1 th 10+1 th
Q2 = 2𝑥( ) =2𝑥( ) = (5.5)th observation = 5𝑡ℎ + 0.5(6𝑡ℎ − 5𝑡ℎ) = 20 +
4 4
0.5(20 − 20) = 20.
n+1 10+1
Q3 = 3𝑥( 4 )th = 3𝑥( 4 )th = (8.25)th observation = 8th + 0.25x(9th - 8th) =
22+0.25x(22-22)= 22

Calculation of quartiles for grouped data

 For the grouped data, the computations of the three quartiles can be done as
follows:
𝑖𝑥𝑛 𝑖𝑥𝑛
 Calculate and search the minimum lcf which is ≥ , 𝑓𝑜𝑟 𝑖 = 1, 2, 3.
4 4
The class corresponding to this lcf is called the ith quartile class. This is the class where Qi
lies.
The unique value of the ith quartile (Qi) is then calculated by the formula

By Admasu M. Lecture Notes Page 11

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

𝒊𝒙𝒏
( 𝟒 − 𝒍𝒄𝒇𝒑 ) 𝒙 𝒘
𝐐𝐢 = 𝐥𝐜𝐛𝒒𝒊 + , 𝒇𝒐𝒓 𝒊 = 𝟏, 𝟐, 𝟑.
𝒇𝒒𝒊
𝑊ℎ𝑒𝑟𝑒: lcb𝑞𝑖 − 𝑖𝑠 𝑡ℎ𝑒 𝒍𝒄𝒃 𝑜𝑓 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑞𝑢𝑎𝑟𝑡𝑖𝑙𝑒 𝑐𝑙𝑎𝑠𝑠.
 𝑓𝑞𝑖 − 𝑖𝑠 𝑡ℎ𝑒 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑞𝑢𝑎𝑟𝑡𝑖𝑙𝑒 𝑐𝑙𝑎𝑠𝑠.
 𝑙𝑐𝑓𝑝 𝑖𝑠 𝑡ℎ𝑒 𝑙𝑐𝑓 𝑐𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑡𝑜 𝑡ℎ𝑒 𝑐𝑙𝑎𝑠𝑠 𝒊𝒎𝒎𝒆𝒅𝒊𝒂𝒕𝒆𝒍𝒚 𝒑𝒓𝒆𝒄𝒆𝒅𝒊𝒏𝒈 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑞𝑢𝑎𝑟𝑡𝑖𝑙𝑒 𝑐𝑙𝑎𝑠𝑠.
𝑁𝑜𝑡𝑒: 𝑄2 = 𝑚𝑒𝑑𝑖𝑎𝑛

2. Percentiles (P)
Percentiles are 99 points, which divide a given ordered data into 100 equal parts. These
𝑚 𝑥 (𝑛 + 1)𝑡ℎ
𝑃𝑚 = , 𝑚 = 1, 2, … ,99. → 𝑓𝑜𝑟 𝑟𝑎𝑤 𝑑𝑎𝑡𝑎.
100

Calculation of percentiles for grouped data

For the grouped data, the computations of the 99 percentiles can be done as follows:
𝑚𝑥𝑛 𝑚𝑥𝑛
 Calculate 100 and search the minimum lcf which is ≥ 100 , 𝑓𝑜𝑟 𝑚 =
1, 2, … ,99.
The class corresponding to this lcf is called the mth percentile class. This is the class where
Pm lies.
The unique value of the mth percentile (Pm)) is then calculated by the formula
𝒎𝒙𝒏
( 𝟏𝟎𝟎 − 𝒍𝒄𝒇𝒑 ) 𝒙 𝒘
𝐩𝐦 = 𝐥𝐜𝐛𝒑𝒎 + , 𝒇𝒐𝒓 𝒎 = 𝟏, 𝟐, … , 𝟗𝟗.
𝒇𝒑𝒎
𝑊ℎ𝑒𝑟𝑒: 𝑙𝑐𝑏𝑝𝑚 , 𝑓𝑝𝑚 𝑎𝑛𝑑 𝑙𝑐𝑓𝑝 𝑤𝑖𝑙𝑙 ℎ𝑎𝑣𝑒 𝑎 𝑠𝑖𝑚𝑖𝑙𝑎𝑟 𝑖𝑛𝑡𝑒𝑟𝑝𝑟𝑒𝑡𝑎𝑡𝑖𝑜𝑛 𝑎𝑠 𝑖𝑛 𝑞𝑢𝑎𝑟𝑡𝑖𝑙𝑒𝑠.
3. Deciles (D)
Deciles are the nine points, which divide the given ordered data into 10 equal parts.
𝑘 𝑥 (𝑛 + 1)𝑡ℎ
𝐷𝑘 = , 𝑘 = 1, 2, … ,9.
10
For the grouped data, the computations of the 9 deciles can be done as follows:
𝑘𝑥𝑛 𝑘𝑥𝑛
 Calculate and search the minimum lcf which is ≥ , 𝑘 = 1, 2, … ,9.
10 10
th
The class corresponding to this lcf is called the k decile class. This is the class where Dk
lies.
The unique value of the kth decile (𝐷𝑘 ) is calculated by the formula
𝒌𝒙𝒏
( 𝟏𝟎 − 𝒍𝒄𝒇𝒑 ) 𝒙 𝒘
𝐃𝐤 = 𝐥𝐜𝐛𝑫𝒌 + , 𝒇𝒐𝒓 𝒌 = 𝟏, 𝟐, … , 𝟗.
𝒇𝑫𝒌
𝑊ℎ𝑒𝑟𝑒: 𝑙𝑐𝑏𝐷𝑘 , 𝑓𝐷𝑘 𝑎𝑛𝑑 𝑙𝑐𝑓𝑝 𝑤𝑖𝑙𝑙 ℎ𝑎𝑣𝑒 𝑎 𝑠𝑖𝑚𝑖𝑙𝑎𝑟 𝑖𝑛𝑡𝑒𝑟𝑝𝑟𝑒𝑡𝑎𝑡𝑖𝑜𝑛 𝑎𝑠 𝑖𝑛 𝑞𝑢𝑎𝑟𝑡𝑖𝑙𝑒𝑠 𝑎𝑛𝑑 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒𝑠.
Note that: 𝑚𝑒𝑑𝑖𝑎𝑛 = Q2 = D5 = P50 and
𝐷1 , 𝐷2 , … , 𝐷9 𝑐𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑 𝑡𝑜 𝑃10 , 𝑃20 , … , 𝑃90
𝑄1 , 𝑄2 𝑎𝑛𝑑 𝑄3 𝑐𝑜𝑟𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑡𝑜 𝑃25 , 𝑃50 𝑎𝑛𝑑 𝑃75
Example: For the following FD data , find
a) 𝑄1 , 𝑄2 𝑎𝑛𝑑 𝑄3 b) 𝑃25 , 𝑃30 , 𝑃50 𝑎𝑛𝑑 𝑃75 c) 𝐷1 , 𝐷2 , 𝐷3 𝑎𝑛𝑑 𝐷5

By Admasu M. Lecture Notes Page 12

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

interval 21 − 22 23 − 24 25 − 26 27 − 28 29 − 30
F 10 22 20 14 14

Solution: First find the lcf table

interval 21 − 22 23 − 24 25 − 26 27 − 28 29 − 30 total
F 10 22 20 14 14 80
Lcf 10 32 52 66 80
𝑛 80
a) 𝑄1 =? = = 20 . Thus, the minimum lcf just ≥ 20 is 32 so the class
4 4
corresponding to this 𝑙𝑐𝑓 𝑖𝑠 23 − 24, is the first quartile class. lcb𝑞1 = 22.5, 𝑤 = 2, 𝑓𝑞1 =
22, 𝑙𝑐𝑓𝑝 = 10.
𝑛
( −𝑙𝑐𝑓𝑝 ) 𝑥 𝑤 (20−10)𝑥2
4
Q1 = lcb𝑞1 + = 22.5 + = 23.41.
𝑓𝑞1 22
2𝑛 160
𝑄2 =? = = 40. Thus, the minimum lcf just ≥ 40 is 52 so the class corresponding
4 4
to this 𝑙𝑐𝑓 𝑖𝑠 25 − 26, is the second quartile class. lcb𝑞2 = 24.5, 𝑤 = 2, 𝑓𝑞2 = 20, 𝑙𝑐𝑓𝑝 =
2𝑥𝑛
( −𝑙𝑐𝑓𝑝 ) 𝑥 𝑤 (40−32)𝑥2
4
32. Q2 = lcb𝑞2 + = 24.5 + = 25.3.
𝑓𝑞2 20
𝑄3 = 27.64.
25𝑥𝑛 25𝑥80
b) 𝑃25 =? = = 20. Thus, the minimum lcf just ≥ 20 is 32 so the class
100 100
corresponding to this 𝑙𝑐𝑓 𝑖𝑠 23 − 24, is the 25th percentile class.
𝑇ℎ𝑢𝑠, lcb𝑝25 = 22.5, 𝑤 = 2, 𝑓𝑝25 = 22, 𝑙𝑐𝑓𝑝 = 10.
25𝑥𝑛
( 100 − 𝑙𝑐𝑓𝑝 ) 𝑥 𝑤 (20 − 10) 𝑥 2
p25 = lcb𝑝25 + = 22.5 + = 23.41.
𝑓𝑝25 22
p20 = 23.045.
p30 = 23.77.
p50 = 25.3.
p75 = 27.64.
1𝑥𝑛 80
C) 𝐷1 =? = 10 = 8. Thus, the minimum lcf just ≥ 8 is 10 so the class corresponding
10
to this 𝑙𝑐𝑓 𝑖𝑠 21 − 22, is the first decile class. 𝑇ℎ𝑢𝑠, lcb𝐷1 = 20.5, 𝑤 = 2, 𝑓𝐷1 =
10, 𝑙𝑐𝑓𝑝 = 0.
1𝑥𝑛
( 10 − 𝑙𝑐𝑓𝑝 ) 𝑥 𝑤 (8 − 0) 𝑥 2
D1 = lcb𝐷1 + = 20.5 + = 22.1 =≫ 𝑢𝑝𝑡𝑜 𝑐𝑙𝑎𝑠𝑠𝑏𝑜𝑢𝑛𝑑𝑎𝑟𝑦.
𝑓𝐷1 10
D2 = 23.045.
D3 = 23.77.
D5 = 25.3.
:. Q1 = P25 , Q2 = P50 and Q3 = P75 D1 = P10, D2 = P20, D3 = P30 and D5 = P50
and median = Q2 = D5 = P50

By Admasu M. Lecture Notes Page 13

 MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

2.4. Measures of variation (dispersion)

Measures of central tendency locate the center of the distribution. But they do not tell how
individual observations are scattered on either side of the center. The spread of observations
around the center is known as dispersion or variability. In other words; the degree to which
numerical data tend to spread about an average value is called dispersion or variation of the
data.
 Small dispersion indicates high uniformity of the observation while larger dispersion
indicates less uniformity.
Types of Measures of Dispersion
The most commonly used measures of dispersions are:
1. The Range (R)
The Range is the difference b/n the highest and the smallest observation. That is;
𝑋𝑚𝑎𝑥 − 𝑋𝑚𝑖𝑛 → 𝑓𝑜𝑟 𝑟𝑎𝑤 𝑎𝑛𝑑 𝑢𝑛𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎.
𝑅={
𝑈𝐶𝐿𝑙𝑎𝑠𝑡 − 𝐿𝐶𝐿𝑓𝑖𝑟𝑠𝑡 → 𝑓𝑜𝑟 𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎.
It is a quick and dirty measure of variability. Because of the range is greatly affected by
extreme values, it may give a distorted picture of the scores.
Range is a measure of absolute dispersion and as such cannot be used for comparing
variability of two distributions expressed in different units.

2. The Quartile Deviation (Q.D) and Coefficient of Quartile Deviation (C.Q.D)

The inter quartile range (IQR) is the difference between the third and the first quartiles of a
set of items. It is not affected by extreme values. So it is a good indicator of the absolute
variability.
Quartile deviation (semi-inter-quartile range): is defined as half of the inter quartile range.
𝑄3 − 𝑄1 𝐼𝑄𝑅
𝑄𝑢𝑎𝑟𝑡𝑖𝑙𝑒 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 (𝑄. 𝐷) = = .
2 2
 It is an absolute measure of variation.
 If quartile deviation is to be used for comparing the variability of two series, then it is
necessary to convert the absolute measure to a coefficient of quartile deviation.
𝑄 −𝑄
𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑄𝑢𝑎𝑟𝑡𝑖𝑙𝑒 𝐷𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 (𝐶. 𝑄. 𝐷) = 𝑄3 +𝑄1 .
1 3

Characteristics of quartile deviation

1. The size of the quartile deviation gives an indication about uniformity. If Q.D is
small, it denotes large uniformity. Thus a coefficient of quartile deviation is used for
comparing uniformity or variation in different distributions.
2. Quartile deviation is not a measure of dispersion in the sense that it does not show the
scatter around on average, but only a distance on scale. Consequently quartile
deviation is regarded as a measure of partition.

By Admasu M. Lecture Notes Page 14

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

3. It can be computed when the distribution has open-ended classes.

Limitation of quartile deviation: It does not satisfy any other test of a good measure of
variation.
Example: For the following frequency distribution, find
a. Inter-quartile range. b. Quartile deviation.
b. Coefficient of quartile deviation. d. Test whether the distribution is symmetric or
not.

CI 21-22 23-24 25-26 27-28 29-30 Total

f 10 22 20 14 14 80
lcf 10 32 52 66 80

𝑄3 −𝑄1 𝐼𝑄𝑅 4.23

a. 𝐼. 𝑄. 𝑅 = 𝑄3 − 𝑄1 = 27.64 − 23.41 = 4.23. 𝑏. 𝑄. 𝐷 = = = =
2 2 2

2.115.
𝑄 −𝑄 4.23
c. 𝐶. 𝑄. 𝐷 = 𝑄3 +𝑄1 = 23.41+27.64 = 0.083
1 3

d. If the distribution is symmetrical

𝑚𝑒𝑑𝑖𝑎𝑛 − 𝑄1 = 𝑄3 – 𝑚𝑒𝑑𝑖𝑎𝑛 , 𝑏𝑢𝑡 𝑚𝑒𝑑𝑖𝑎𝑛 = 𝑄2 = 25.3.
𝐼𝑛 𝑜𝑢𝑟 𝑐𝑎𝑠𝑒: 𝑚𝑒𝑑𝑖𝑎𝑛 − 𝑄1 = 𝑄3 – 𝑚𝑒𝑑𝑖𝑎𝑛 =≫ 1.89 = 2.34 → 𝑓𝑎𝑙𝑠𝑒
Therefore, the distribution is not symmetrical.
3. Mean Deviation (Average deviation) and Coefficient of Mean Deviation
The mean deviation measures the scatter of individual observations around a central value
(usually the mean or the median) of a distribution. So, the mean deviation is defined as the
arithmetic mean of positive deviations of each observation from either mean or the median of
a distribution.
 If the deviations are taken from the mean then it is called mean deviation from the
mean. If on the other hand, the deviations are taken from the median we call it mean
deviation from the median.
∑𝑛𝑖=1|𝑋𝑖 − 𝑋̅|
𝑖. 𝑒. 𝑀𝑒𝑎𝑛 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑚𝑒𝑎𝑛 ( 𝑀𝐷𝑥̅ ) =
𝑛
∑𝑛𝑖=1|𝑋𝑖 − 𝑋̃|
𝑀𝑒𝑎𝑛 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑚𝑒𝑑𝑖𝑎𝑛 (𝑀𝐷𝑥̃ ) =
𝑛
Example: The weight of a sample of six students from a class (in Kg) is given below:
53, 56, 57, 59, 63, and 66.
i. What is the mean deviation from the mean?
ii. What is the mean deviation from the median?
53+56+⋯+66
Solution: 𝑋̅ = = 59 , and 𝑋̃ = 58.
6

By Admasu M. Lecture Notes Page 15

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

∑𝑛 ̅
𝑖=1|𝑋𝑖 − 𝑋 | |53− 59|+|56− 59|+⋯+|66− 59|
𝑀𝐷𝑥̅ = = = 3.67.
𝑛 6
∑𝑛 ̃
𝑖=1|𝑋𝑖 − 𝑋 | |53− 58|+|56− 58|+⋯+|66− 58|
𝑀𝐷𝑥̃ = = = 3.67.
𝑛 6

Mean deviation for grouped data

For grouped data, the mean deviation from the mean and median are given by
∑𝑘 ̅
𝑖=1 𝑓𝑖 |𝑋𝑖 − 𝑋 |
𝑀𝐷𝑥̅ = ∑ 𝑓𝑖
𝑘 } → For ungrouped frequency distribution.
∑𝑖=1 𝑓𝑖 |𝑋𝑖 − 𝑋̃ |
𝑀𝐷𝑥̃ = ∑ 𝑓𝑖

∑𝑘 ̅
𝑖=1 𝑓𝑖 |𝑚𝑖 − 𝑋|
𝑀𝐷𝑥̅ = ∑ 𝑓𝑖
𝑘
} → For grouped frequency distribution.
∑𝑖=1 𝑓𝑖 |𝑚𝑖 − 𝑋̃|
𝑀𝐷𝑥̃ = ∑ 𝑓𝑖

The coefficient of mean deviation from the mean and from the median are given by:
𝑀𝐷𝑥̅
𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑀𝐷𝑥̅ =
𝑋̅
𝑀𝐷𝑥̃
𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑀𝐷𝑥̃ =
𝑋̃
Example: Find the coefficient of mean deviation from the mean and from the median for the
weight of six students in previous example.

𝑀𝐷𝑥̅ 3.67 𝑘𝑔
Solution: Coefficient of 𝑀𝐷𝑥̅ = = = 0.0622.
𝑋̅ 59 𝑘𝑔
𝑀𝐷𝑥̃ 3.67 𝑘𝑔
Coefficient of 𝑀𝐷𝑥̃ = = = 0.0633.
𝑋̃ 58 𝑘𝑔

Ex: Find the mean deviation and coefficient of mean deviation from the mean and from the
median for the data given below, test results of 10 students collected (out of 20).

Test result (CI) 1-5 6-10 11-15 16-20

No of students (f) 4 1 2 3

Limitation: Because of absolute value, mean deviation ignores the algebraic signs of the
deviations and it leads to a series difficulties in inference.
4. Variance and Standard Deviation
Variance: is the average of the squares of the deviations taken from the mean.
Suppose that 𝑥1 , 𝑥2 , … , 𝑥𝑁 be the set of observations on N populations. Then,
∑𝑁
𝑖=1(𝑥𝑖 − 𝜇)
2 ∑𝑁 2
𝑖=1 𝑥𝑖 − 𝑁𝜇
2
𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 = 𝜎 2 = = . → 𝑓𝑜𝑟 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛.
𝑁 𝑁
∑𝑛
𝑖=1(𝑥𝑖 − 𝑥̅ )2 ∑𝑛
𝑖=1 𝑥𝑖
2 − 𝑛𝑥̅ 2
𝑆𝑎𝑚𝑝𝑙𝑒 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 = 𝑠 2 = = . → 𝑓𝑜𝑟 𝑠𝑎𝑚𝑝𝑙𝑒.
𝑛−1 𝑛−1
In general, the sample variance is computed by:

By Admasu M. Lecture Notes Page 16

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

∑𝑛𝑖=1(𝑥𝑖 − 𝑥̅ )2 ∑𝑛𝑖=1 𝑥𝑖 2 − 𝑛𝑥̅ 2

= . → 𝑓𝑜𝑟 𝑟𝑎𝑤 𝑑𝑎𝑡𝑎.
𝑛−1 𝑛−1
∑𝑘𝑖=1 𝑓𝑖 (𝑥𝑖 − 𝑥̅ )2 ∑𝑘𝑖=1 𝑓𝑖 𝑥𝑖 2 − 𝑛𝑥̅ 2
𝑠2 = = . → 𝑓𝑜𝑟 𝑢𝑛𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎.
∑𝑘𝑖=1 𝑓𝑖 − 1 𝑛−1
∑𝑘𝑖=1 𝑓𝑖 (𝑚𝑖 − 𝑥̅ )2 ∑𝑘𝑖=1 𝑓𝑖 𝑚𝑖 2 − 𝑛𝑥̅ 2
= . → 𝑓𝑜𝑟 𝑔𝑟𝑜𝑢𝑝𝑒𝑑 𝑑𝑎𝑡𝑎.
{ ∑𝑘𝑖=1 𝑓𝑖 − 1 𝑛−1
Standard Deviation: it is the square root of variance. Its advantage over the variance is that it
is in the same units as the variable under the consideration. It is a measure of the average
variation in a set of data. It is a measure of how far, on the average, an individual
measurements is from the mean. 𝑆. 𝑑 = √𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 = √𝑆 2 = 𝑆.
Example 1: Compute the variance for the sample: 5, 14, 2, 2 and 17.
Solution: 𝑛 = 5 , ∑𝑛𝑖=1 𝑥𝑖 = 40, 𝑥̅ = 8 , ∑𝑛𝑖=1 𝑥𝑖 2 = 518 .
2
∑𝑛𝑖=1 𝑥𝑖 2 − 𝑛𝑥̅ 2 518 − 5 𝑥 82
𝑠 = = = 49.5. , 𝑆 = √49.5
𝑛−1 5−1
= 7.04.
Example 2: Suppose the data given below indicates time in minute required for a laboratory
experiment to compute a certain laboratory test. Calculate the mean, variance and standard
deviation for the following data.
𝒙𝒊 32 36 40 44 48 Total
𝒇𝒊 2 5 8 4 1 20
𝒇𝒊 𝒙𝒊 64 180 320 176 48 788
𝟐 2048 6480 12800 7744 2304 31376
𝒇𝒊 𝒙𝒊
2
∑𝑛
𝑖=1 𝒇𝒊 𝒙𝒊 788 ∑𝑛
𝑖=1 𝒇𝒊 𝑥𝑖 − 𝑛𝑥̅
2 31376−20 𝑥 (39.4)2
𝑥̅ = = = 39.4 , 𝑠 2 = = = 17.31. , 𝑆 =
𝑛 20 𝑛−1 19
√17.31 = 4.16.
Properties of Variance
1. The variance is always non-negative ( 𝑠 2 ≥ 0).
2. If every element of the data is multiplied by a constant "c", then the new variance
𝑠 2 𝑛𝑒𝑤 = 𝑐 2 𝑥 𝑠 2 𝑜𝑙𝑑 .
3. When a constant is added to all elements of the data, then the variance does not
change.
4. The variance of a constant (c) measured in n times is zero. i.e. (var(c) = 0).
Exercise: Verify the above properties.
Uses of the Variance and Standard Deviation
1. They can be used to determine the spread of the data. If the variance or S.D is large,
then the data are more dispersed.
2. They are used to measure the consistency of a variable.
3. They are used quit often in inferential statistics.
5. Coefficient of Variation (C.V)
Whenever the two groups have the same units of measurement, the variance and S.D for each
can be compared directly. A statistics that allows one to compare two groups when the units
of measurement are different is called coefficient of variation. It is computed by:

By Admasu M. Lecture Notes Page 17

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

𝜎
𝐶. 𝑉 = 𝑥 100% → 𝑓𝑜𝑟 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛.
𝜇
𝑆
𝐶. 𝑉 = 𝑥 100% → 𝑓𝑜𝑟 𝑠𝑎𝑚𝑝𝑙𝑒.
𝑥̅
Example: The following data refers to the hemoglobin level for 5 males and 5 female
students. In which case, the hemoglobin level has high variability (less consistency).

For males (xi) 13 13.8 14.6 15.6 17

For females (xi) 12 12.5 13.8 14.6 15.6

13+13.8+14.6+15.6+17 74 12+12.5+13.8+14.6+15.6
Solution: 𝑥̅𝑚𝑎𝑙𝑒 = = = 14.8 , 𝑥̅𝑓𝑒𝑚𝑎𝑙𝑒 = =
5 5 5
68
= 13.7.
5

∑𝑛 2
𝑖=1 𝑥𝑖 − 𝑛𝑥̅
2
𝑠 2 𝑚𝑎𝑙𝑒𝑠 = = 2.44. , 𝑆𝑚𝑎𝑙𝑒𝑠 = √2.44 = 1.56205.,
𝑛−1

∑𝑛𝑖=1 𝑥𝑖 2 − 𝑛𝑥̅ 2
𝑠 2𝑓𝑒𝑚𝑎𝑙𝑒𝑠 = = 2.19. , 𝑆𝑓𝑒𝑚𝑎𝑙𝑒𝑠 = √2.19 = 1.479865.
𝑛−1
𝑆𝑚𝑎𝑙𝑒𝑠 1.56205
𝐶. 𝑉𝑚𝑎𝑙𝑒𝑠 = 𝑥 100% = 𝑥100% = 𝟏𝟎. 𝟓𝟔%,
𝑥̅ 𝑚𝑎𝑙𝑒 14.8

𝑆𝑓𝑒𝑚𝑎𝑙𝑒𝑠 1.479865
𝐶. 𝑉𝑓𝑒𝑚𝑎𝑙𝑒𝑠 = 𝑥 100% = 𝑥100% = 𝟏𝟎. 𝟖%.
𝑥̅𝑓𝑒𝑚𝑎𝑙𝑒 13.7

Therefore, the variability in hemoglobin level is higher for females than for males.

6. Standard Scores (Z-Scores)

 It is used for describing the relative position of a single score in the entire set of data
in terms of the mean and standard deviation.
 It is used to compare two observations coming from different groups.
 If X is a measurement (an observation) from a distribution with mean 𝑥̅ and
standard deviation S, then its value in standard units is
𝑋−𝜇
𝑍= 𝜎 → 𝑓𝑜𝑟 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛.
𝑥 − 𝑥̅
𝑍= → 𝑓𝑜𝑟 𝑠𝑎𝑚𝑝𝑙𝑒.
𝑆
 Z gives the number of standard deviation a particular observation lie
above or below the mean.
 A positive Z-score indicates that the observation is above the mean.
 A negative Z-score indicates that the observation is below the mean.
Example: Two sections were given an examination on a certain course. For section 1, the
average mark (score) was 72 with standard deviation of 6 and for section 2, the average mark
(score) was 85 with standard deviation of 7. If student A from section 1 scored 84 and student
B from section 2 scored 90, then who perform a better relative to the group?
𝑥−𝑥̅ 84−72
Solution: 𝑍 𝑠𝑐𝑜𝑟𝑒 𝑓𝑜𝑟 𝐴 𝑖𝑠 𝑍 = = = 2.
𝑆 6

By Admasu M. Lecture Notes Page 18

MTU Stat Dept Probability and Statistics for engineer's Chapter - 2

𝑥−𝑥̅ 90−85
𝑍 𝑠𝑐𝑜𝑟𝑒 𝑓𝑜𝑟 𝐵 𝑖𝑠 𝑍 = = = 0.71.
𝑆 7
Since ZA > ZB i.e. 2 > 0.71, student A performed better relative to his group than student B.
Therefore, student A has performed better relative to his group because the score's of student
A is two standard deviation above the mean score of section 1 while the score of student B is
only 0.71 standard deviation above the mean score of students in section 2.

By Admasu M. Lecture Notes Page 19