Describing Group

Performance
Quartiles, Percentiles, Deciles
Measures of A measure of position determines the position of a single
value in relation to other values or a population data set.
Position
 Quartiles
Deciles Decile Rank Percentile Percentage
100% 𝑸𝟒
10th decile 10 100%
4th Quartile
9th decile 9 90th percentile 90%
75% 𝑸𝟑 8th decile 8 80th percentile 80%
3rd Quartile 7th decile 7 70th percentile 70%
6th decile 6 60th percentile 60%
50% 𝑸𝟐
5th decile 5 50th percentile 50%
2nd Quartile
4th decile 4 40th percentile 40%
25% 𝑸𝟏 3rd decile 3 30th percentile 30%
1st Quartile 2nd decile 2 20th percentile 20%
1st decile 1 10th percentile 10%
Quartiles

A set of observation may be divided

into four equal parts called quartile.
The upper quartile (Q3)
known as the third quartile is the
value of the variable below which
75% of the cases lie. The third
quartile also corresponds to the 75th
percentile point because it
surpasses 75% of the cases. Hence,
Q3= P75
Quartiles

The lower quartile (Q1)

known as the first quartile, is the
value of the variable below which
25% of the cases lie. It also
corresponds to the 25th percentile
point. Hence, Q1 = P25
Therefore, we can conclude
that the median corresponds to the
second quartile and 50th percentile,
thus, X = Q2 = P50
Formulas in Computing Quartiles in Ungrouped Data

If a data set of values is arranged in ascending order of magnitude, then: Quartiles

100% 𝑸𝟒
❑ The lower quartile (Q1) is the median of the lower half of the data set.
1 4th Quartile
• 𝑄1 = 𝑛 + 1 𝑡ℎ the value where n is the number of data values in
4 𝑸𝟑
the data set 75%

❑ The second quartile (𝑄2 ) is the middle value of the data set. Note: it is equal 3rd Quartile
to the median value of the central tendency 50% 𝑸𝟐
2
• 𝑄2 = 𝑛 + 1 𝑡ℎ the value where n is the number of data values in 2nd Quartile
4
the data set
25% 𝑸𝟏
❑ The upper quartile (𝑄2 ) is the median of the upper half of the data set.
3 1st Quartile
• 𝑄2 = 𝑛 + 1 𝑡ℎ the value where n is the number of data values in
4
the data set
Example:
Find the median, lower quartile, and upper quartile of the
following data set of scores:
18 20 23 20 23 27 24 23 29

Solution:
Arrange the values in ascending order of magnitude:

18 20 20 23 23 23 24 27 29

There are 9 values in the data set.

𝑛=9
 Data set: 18 20 20 23 23 23 24 27 29

2
Median or Second quartile (Q2)= 𝑛 + 1 𝑡ℎ value
4
2
= 9 + 1 𝑡ℎ value
4

2
= 10 𝑡ℎ value
4

= 𝟓𝒕𝒉 value

= 𝟐𝟑

Analysis: 50% of the data falls below 23.

 Data set: 18 20 20 23 23 23 24 27 29

1
lower quartile (Q1)= 𝑛 + 1 𝑡ℎ value
4
1
= 9 + 1 𝑡ℎ value
4
1
= 10 𝑡ℎ value
4

= 𝟐. 𝟓𝒕𝒉 value
average of 2nd and 3rd values

20 + 20
=
2
40
=
2
= 𝟐𝟎 Analysis: 25% of the data falls below 20.
 Data set: 18 20 20 23 23 23 24 27 29

3
Upper quartile (Q3)= 𝑛 + 1 𝑡ℎ value Q3
4

3
= 9 + 1 𝑡ℎ value
4

3
= 10 𝑡ℎ value
4
30
= 𝑡ℎ value
4

= 𝟕. 𝟓𝒕𝒉 value
average of 7th and 8th values
24 + 27
=
2
51
=
2
= 𝟐𝟓. 𝟓

Analysis: 75% of the data falls below 25.5.

Formulas in Computing Quartiles in a Frequency Distribution (Grouped Data)

Computing the value of the first and third quartile is similar in computing the value of the median:

𝑘𝑛
− 𝑐𝑓𝑏
𝑄𝑘 = 𝐿𝑏 + 4 c. i.
𝑓𝑞𝑘

Where:
LB - lower boundary of the kth quartile class
Cfb – cumulative frequency before the kth quartile class
𝑓𝑞𝑘 – frequency of the kth quartile class

c.i. – class size

Formulas of the quartiles when k = 1, 2, 3

1𝑛
− 𝑐𝑓𝑏
𝑄1 = 𝐿𝑏 + 4 c. i.
𝑓𝑞1

2𝑛
− 𝑐𝑓𝑏
𝑄2 = 𝐿𝑏 + 4 c. i.
𝑓𝑞2

3𝑛
− 𝑐𝑓𝑏
𝑄3 = 𝐿𝑏 + 4 c. i.
𝑓𝑞3
Example: Consider the test scores of 50 students in Basic Statistics students in a certain
college. Find Q1, Q2, and Q3

Class Interval Frequency (f) Less Than Cumulative

(ci) Frequency (<cf)
12-16 6 6
17-21 5 11
22-26 10 21
27-31 7 28
32-36 6 34
37-41 9 43
42-46 6 49
47-51 1 50
N=50
Solving for 𝑄1 :
1𝑛
− 𝑐𝑓𝑏
1𝑛
Solve for first to determine 𝑄1 = 𝐿𝑏 + 4 c. i.
Class Interval (ci) Frequency (f) <cf 4 𝑓𝑞1
the quartile class
12-16 6 6
50
17-21 5 11 − 11
1𝑛
=
(1)50
= 21.5 + 4 5
22-26 10 21
4 4 10
= 12.5
27-31 7 28 12.5 − 11
= 21.5 + 5
32-36 6 34 𝑄1 Class = 22 – 26 10
37-41 9 43 𝐿𝑏 = 21.5
1.5
42-46 6 49 𝑐𝑓𝑏 = 11 = 21.5 + 5
10
47-51 1 50 𝑓𝑞1 = 10
7.5
N=50 = 21.5 +
𝑐. 𝑖 = 5 10

= 𝟐𝟐. 𝟐𝟓 or 22

Analysis: approximately 25% of the students’

scores fall below the score 22.
Solving for 𝑄2 or Median:
2𝑛
− 𝑐𝑓𝑏
Class Interval (ci) Frequency (f) <cf
2𝑛
Solve for first to determine 𝑄2 = 𝐿𝑏 + 4 c. i.
4 𝑓𝑞2
the quartile class
12-16 6 6
17-21 5 11 25 − 21
2𝑛 (2)50 = 26.5 + 5
= 7
22-26 10 21 4 4

27-31 7 28 = 25 4
= 26.5 + 5
7
32-36 6 34 𝑄2 Class = 27 – 31
37-41 9 43 20
𝐿𝑏 = 26.5 = 26.5 +
7
42-46 6 49 𝑐𝑓𝑏 = 21
47-51 1 50 = 𝟐𝟗. 𝟒 or 29
𝑓𝑞2 = 7
N=50
𝑐. 𝑖 = 5
Analysis: approximately 50%
of the students’ scores fall
below the score 29.
Solving for 𝑄3 :
3𝑛
− 𝑐𝑓𝑏
Class Interval (ci) Frequency (f) <cf
3𝑛
Solve for first to determine 𝑄3 = 𝐿𝑏 + 4 c. i.
4 𝑓𝑞3
the quartile class
12-16 6 6
37.5 − 34
17-21 5 11 3𝑛 (3)50 = 36.5 + 5
= 9
22-26 10 21 4 4

27-31 7 28 = 37.5 3.5

= 36.5 + 5
32-36 6 34 9
𝑄3 Class = 37 – 41
37-41 9 43 𝐿𝑏 = 36.5 17.5
= 36.5 +
42-46 6 49 𝑐𝑓𝑏 = 34 9
47-51 1 50
𝑓𝑞3 = 9
N=50 = 𝟑𝟖. 𝟒𝟒 or 38
𝑐. 𝑖 = 5
Analysis: approximately 75% of
the students’ scores fall below
the score 38.
Percentiles
One way of assessing performance is to
determine the position of one with respect to others.
Let us say Rosalina got a score of 72 in a 100-
item test and is ranked 8th in a class of 145 students, it
follows that there are 145 − 8 = 137 students below
her in rank. If we divide this number by the total
number of students, we get 137/145 or 91.5% or 92%.
This means that 92% of the class ranked below or got
scores below 72.
We can say therefore that the percentile rank
of Rosalina in the class is 92. This means also that 92
out of every 100 students got scores below her score.
Conversely, this could also mean that 8% of the
students have higher scores than Rosalina. Since
Rosalina’s score is 72, we can further say that the 92th
percentile is 72.
Percentiles
Refers to those values that
divide a distribution into one hundred
equal parts.
Percentile rank n tells how many
percent of the cases got below the rank
position.
Percentile point Pn is the score
or value that corresponds to the given
percentile rank. Thus, in the example,
P92= 72.
 Percentiles for the ungrouped data:
To calculate percentiles (a measure of the relative
standing of an observation) for the ungrouped data, adopt
the following procedure

1. Order the observation

𝑚(𝑛+1)
2. For the mth percentile, determine the product .
100
𝑚(𝑛+1)
If is not an integer, round it up and find the
100
corresponding ordered value and if m.n100 is an
integer, say k, then calculate the mean of the Kth
and (k+1)th ordered observations.
Example: The following are scores of first year students in their major examinations in Math
Enhancement. Find the 10th and 95th percentiles.

91, 89, 88, 87, 89, 91, 87, 92, 90, 98, 95, 97, 96, 100, 101, 96, 98, 99, 98, 100, 102, 99, 101,
105, 103, 107, 105, 106, 107, 112.

Solution:
Step 1:Order the observation
87 98
87 99
88 99
89 100
89 100
90
91
101
101
𝑛 = 30
91 102
92 103
95 105
96 105
96 106
97 107
98 107
98 112
Step 2: Solve for the mth percentile 87 98
87 99
𝑚 × (𝑛 + 1) 88 99
𝑃𝑚 =
100 89 100
89 100
Solving for 𝑃10 :
90 101
10 × (30 + 1) 91 101
𝑃10 = 91 102
100
92 103
10 × 31 95 105
= 96 105
100 96 106
97 107
= 𝟑. 𝟏 𝐭𝐡 𝐯𝐚𝐥𝐮𝐞 98 107
98 112
𝟑. 𝟏 𝐭𝐡 𝐯𝐚𝐥𝐮𝐞 = 3𝑟𝑑 𝑣𝑎𝑙𝑢𝑒 + 0.1(4𝑡ℎ − 3𝑟𝑑 𝑣𝑎𝑙𝑢𝑒)

= 88 + 0.1(89 − 88)

= 88 + 0.1(1)
So 𝑃10 or 10th percentile is the 3.1th observation of the sorted data, which is 88.1.
= 88.1 Which means that 10% of the observation of the data set are less than 88.1.
 Solving for 𝑃95 :

95 × (30 + 1)
𝑃95 = 87 98
100 87 99
2945 88 99
= 89 100
100
89 100
=29.45th value 90 101
91 101
91 102
29.45th value= 29𝑡ℎ 𝑣𝑎𝑙𝑢𝑒 + 0.45(30𝑡ℎ 𝑣𝑎𝑙𝑢𝑒 − 29𝑡ℎ 𝑣𝑎𝑙𝑢𝑒) 92 103
95 105
= 107 + 0.45(112 − 107) 105
96
= 107 + 0.45(5) 96 106
97 107
= 107 + 2.25 98 107
98 112
= 109.25

*So 𝑃95 or 95th percentile is the 29.45th observation of the sorted data, which is 109.25. Which
means that 95% of the observation of the data set are less than 109.25.
What if… 87
87
98
99
88 99
𝑃𝑘 =29.5th value 89 100
89 100
90 101
Simply get the average of the 29th and 30th value: 91 101
91
107 + 112 102
= 92
2 103
95
105
219 96
105
= 96
2 106
97
107
= 109.5 98
107
98
112
Computation of Percentiles for Grouped Data

𝐾𝑛
− 𝑐𝑓𝑏
𝑃𝐾 = 𝐿𝑏 + 100 c. i.
𝑓𝑃𝐾

where : thus, when k = 70, the formula is

Lb = lower boundary of the kth percentile class
cfb = cumulative frequency before the kth percentile 70𝑛
− 𝑐𝑓𝑏
class 𝑃70 = 𝐿𝑏 + 4 c. i.
fpk = frequency of the kth percentile class 𝑓𝑃70
ci =class size
Example: Consider test scores of 50 students in Basic Statistics students in a certain college.
Determine the value of 75th, 90th, and 40th percentiles. Using frequency distribution below.

Class Interval Frequency (f) Less Than Cumulative

(ci) Frequency (<cf)
12-16 6 6
17-21 5 11
22-26 10 21
27-31 7 28
32-36 6 34
37-41 9 43
42-46 6 49
47-51 1 50
N=50
Solving for 75th percentile:
75𝑛
− 𝑐𝑓𝑏
Class Interval (ci) Frequency (f) <cf Solve for
75𝑛
first to determine 𝑃75 = 𝐿𝑏 + 100 c. i.
100
the percentile class
𝑓𝑃𝐾
12-16 6 6
17-21 5 11 𝑘𝑛 75(50) 37.5 − 34
= = 37.5 = 36.5 + 5
100 100 9
22-26 10 21
𝑃75 𝐶𝑙𝑎𝑠𝑠 = 37 − 41
27-31 7 28 3.5
= 36.5 + 5
32-36 6 34 𝐿𝑏 = 36.5 9
37-41 9 43 𝑐𝑓𝑏 = 34 3.5
= 36.5 + 5
42-46 6 49 𝑓𝑃75 = 9 9
47-51 1 50 𝑐. 𝑖 = 5 17.5
= 36.5 +
N=50 9

= 𝟑𝟖. 𝟒𝟒 or 38

Analysis: approximately 75% of the

students’ scores fall below the score 38.
Solving for 90th percentile:
90𝑛
− 𝑐𝑓𝑏
Class Interval (ci) Frequency (f) <cf Solve for
90𝑛
first to determine 𝑃90 = 𝐿𝑏 + 100 c. i.
100
the percentile class
𝑓𝑃𝐾
12-16 6 6
17-21 5 11 𝑘𝑛 90(50) 45 − 43
= = 45 = 41.5 + 5
100 100 6
22-26 10 21
𝑃90 𝐶𝑙𝑎𝑠𝑠 = 42 − 46
27-31 7 28 2
= 41.5 + 5
32-36 6 34 𝐿𝑏 = 41.5 6
37-41 9 43 𝑐𝑓𝑏 = 43
10
42-46 6 49 𝑓𝑃90 = 6 = 41.5 +
6
47-51 1 50 𝑐. 𝑖 = 5
N=50 = 𝟒𝟑. 𝟏𝟕

Analysis: approximately 90% of the

students’ scores fall below the score 43.
Solving for 40th percentile:
40𝑛
− 𝑐𝑓𝑏
Class Interval (ci) Frequency (f) <cf Solve for
40𝑛
first to determine 𝑃90 = 𝐿𝑏 + 100 c. i.
100
the percentile class
𝑓𝑃𝐾
12-16 6 6
17-21 5 11 𝑘𝑛 40(50)
= = 20 20 − 11
100 100 = 21.5 + 5
22-26 10 21 10
𝑃40 𝐶𝑙𝑎𝑠𝑠 = 22 − 26
27-31 7 28 9
= 21.5 + 5
32-36 6 34 𝐿𝑏 = 21.5 10
37-41 9 43 𝑐𝑓𝑏 = 11
45
42-46 6 49 𝑓𝑃40 = 10 = 21.5 +
10
47-51 1 50 𝑐. 𝑖 = 5
N=50 = 𝟐𝟔

Analysis: approximately 40% of the

students’ scores fall below the score 26.
 Deciles
When the distribution is divided
into 10 equal parts, each part is called
decile. A decile point Dk is a value of
the variables below which a certain
percentage of the class fall, the
percentage being taken in units of 10.

For example, the 6th decile point

is the score which surpasses 60%, of
the individuals in the distribution. This
means that D6= P60
 The computing formula for grouped data is patterned after the formula of median,
quartiles or percentiles.

𝑘𝑛
− 𝑐𝑓𝑏
𝐷𝑘 = 𝐿𝑏 + 10 c. i.
𝑓𝑑𝐾

Where:
Thus, when k=7, the formula is
LB=lower boundary of the kth decile class
cfb = cumulative frequency before the kth decile class 7𝑛
− 𝑐𝑓𝑏
𝐷7 = 𝐿𝑏 + 10 c. i.
fdk = frequency of the kth decile class
𝑓𝑑7
c.i. = class size
Example: Determine the value of the 7thdecile of the given frequency distribution below

Class Interval Frequency (f) Less Than Cumulative

(ci) Frequency (<cf)
12-16 6 6
17-21 5 11
22-26 10 21
27-31 7 28
32-36 6 34
37-41 9 43
42-46 6 49
47-51 1 50
N=50
Solving for 𝐷7
𝑘𝑛
− 𝑐𝑓𝑏
7𝑛 𝐷𝑘 = 𝐿𝑏 + 10 c. i.
Class Interval (ci) Frequency (f) <cf Solve for first to determine 𝑓𝑑𝐾
10
12-16 6 6 the decile class
7𝑛
7(50) − 𝑐𝑓𝑏
17-21 5 11 = 35 𝐷7 = 𝐿𝑏 + 10 c. i.
10 𝑓𝑑7
22-26 10 21
𝐷7 𝐶𝑙𝑎𝑠𝑠 = 37 − 41 35 − 34
27-31 7 28
𝐷7 = 36.5 + 5
32-36 6 34 𝐿𝑏 = 36.5 9
37-41 9 43 𝑐𝑓𝑏 = 34 1
𝐷7 = 36.5 + 5
42-46 6 49 𝑓𝑑7 = 9 9
47-51 1 50 𝑐. 𝑖 = 5 5
𝐷7 = 36.5 +
N=50 9

𝑫𝟕 = 𝟑𝟕. 𝟏

Analysis: approximately 70% of the students’

scores fall below the score 37.1.
Exercise