Scribd
Upload
29 views4 pages

Statistics and Probability

Measures of central tendency describe the central or typical values in a data set and include the mean, median, and mode. The mean is the average and is calculated by adding all values and dividing by the total number of values. The median is the middle value that separates the higher half from the lower half of a data set. The mode is the value that occurs most frequently. Partition values like quartiles further divide a data set into equal parts to help summarize the central location or spread of the values.

Uploaded by

Aditya Messi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
29 views4 pages

Statistics and Probability

Measures of central tendency describe the central or typical values in a data set and include the mean, median, and mode. The mean is the average and is calculated by adding all values and dividing by the total number of values. The median is the middle value that separates the higher half from the lower half of a data set. The mode is the value that occurs most frequently. Partition values like quartiles further divide a data set into equal parts to help summarize the central location or spread of the values.

Uploaded by

Aditya Messi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Measures of Central Tendency

Every data-set has central concentration pattern called central tendency. The
single value that can represent whole set of data is called measure of central
tendency (where whole set of data is supposed to be concentrated).

Requisites of good measure of central tendency

i. It should be rigidly defined.

ii. It should be based on all the observations.

iii. It should be easy to calculate and simple to understand.

iv. It should be suitable for further mathematical treatment.

v. It should be least affected by fluctuation in sampling.

vi. It should not be affected by extreme observations.

1. Arithmetic mean (AM): - It is denoted by 𝑥̅ and given by;


∑𝑥
a. For individual series, 𝑥̅ = , where x=value of variable, n= number
𝑛
of values and Σ=summation sign.

b. For discrete and continuous series (i.e., for frequency distribution) 𝑥̅ =


∑ 𝑓𝑥
,where x=value of variable (mid value for continuous classes),
𝑁
f=frequency of x and N= Σ f= total frequency

∑ 𝑓𝑢
c. For change of origin and/or scale (coding) method, 𝑥̅ = 𝐴 + ×ℎ
𝑁
Where, A=Assumed mean (Usually mid value of middle class)
𝑥−𝐴
h=Non-zero scale factor (usually class size) and 𝑢 =
ℎ

∑ 𝑤𝑥
d. Weighted AM: - It is given by 𝑥̅𝑤 = ∑𝑤
Where, 𝑤=
𝑊𝑒𝑖𝑔ℎ𝑡 𝑎𝑠𝑠𝑖𝑔𝑛𝑒𝑑 𝑡𝑜 𝑥
e. Combined mean: - For two series of data;

𝑛1 𝑥̅1 + 𝑛2 𝑥̅2
𝑥̅12 =
𝑛1 + 𝑛2
Where, n1=size of first series of data
n2= size of second series of data
𝑥̅1 = 𝐴𝑀 𝑜𝑓 𝑓𝑖𝑟𝑠𝑡 𝑠𝑒𝑟𝑖𝑒𝑠 𝑜𝑓 𝑑𝑎𝑡𝑎 𝑥̅2 = 𝐴𝑀 𝑜𝑓 𝑠𝑒𝑐𝑜𝑛𝑑 𝑠𝑒𝑟𝑖𝑒𝑠 𝑜𝑓 𝑑𝑎𝑡𝑎
Note: Arrange the data as follows and solve the problem

SN Mean Size Total


1. 𝑥̅1 𝑛1 ∑ 𝑥1 = 𝑛1 𝑥̅1
2. 𝑥̅2 𝑛2 ∑ 𝑥2 = 𝑛2 𝑥̅2
3. 𝑥̅12 𝑛1 + 𝑛2 𝑛1 𝑥̅1 + 𝑛2 𝑥̅2

2. Mode (Mo): - Mode is the most repeated item. It is also understood as the
favorite item, fashionable item, customers’ preference or the most frequent
item. For individual and discrete series, mode is determined by inspection. But,
in continuous series with exclusive type classes, modal class is determined by
𝑓1 −𝑓𝑜
inspection and mode is computed by using the formula; 𝑀𝑜 = 𝑙 + ×ℎ
2𝑓1 −𝑓𝑜 −𝑓2
Where, l=lower limit of modal class
f1=Frequency of modal class/maximum frequency
fo=Frequency preceding (अघिल्लो) the modal class
f2=Frequency following(पघिल्लो) the modal class
h=size/width of modal class.
3. Partition values: - Those values which divide whole set of data into a number
of equal parts are called partition values. They are; Median (Md), Quartiles
(Qi), Deciles (Dj) and Percentiles (Pk).
a. Median (Md): - Median is a single value which divide whole set of data
into two equal parts, 50% each.

50% 50%
Smallest Md Largest
item item

b. Quartiles (Qi, i=1,2,3):- Quartiles are the three values which divide
whole set of data into four equal parts/quarters, 25% each. The first
quartile is called lower quartile, the third quartile is called upper
quartile and the second quartile is Median.

S Q1 L
Q2 Q3
25% 25% 25% 25%
Md

c. Deciles (Dj, j=1,2,…..9): - Deciles are the nine values which divide whole
set of data into ten equal parts, 10% each.
10% 10% 10%
S D1 D2 D5 D9 L

a. Percentiles (Pk, k=1,2, … ... 99): - Percentiles are the ninety-nine values
which divide whole set of data into hundred equal parts, 1% each.

S P1 P2 P50 Pk P90 L
1% 1%
k% (100-k) %

Steps in computing Partition values;


- Arrange the data in ascending order of their magnitude
- Compute (less than) cumulative frequencies for frequency distributions (i.e.
discrete and continuous series)
- Then,
For individual and discrete series;
𝑛+1 𝑛𝑑
[ ] 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙 𝑠𝑒𝑟𝑖𝑒𝑠
2
a. 𝑀𝑑 = { 𝑁+1
[ ]𝑛𝑑 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑑𝑖𝑠𝑐𝑟𝑒𝑡𝑒 𝑠𝑒𝑟𝑖𝑒𝑠.
2
𝑛+1 𝑡ℎ
𝑖[ ] 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙 𝑠𝑒𝑟𝑖𝑒𝑠
4
b. 𝑄𝑖 = { 𝑁+1 𝑡ℎ
𝑖[ ] 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑑𝑖𝑠𝑐𝑟𝑒𝑡𝑒 𝑠𝑒𝑟𝑖𝑒𝑠.
4
i = 1, 2, 3
𝑛+1 𝑡ℎ
𝑗[ ] 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙 𝑠𝑒𝑟𝑖𝑒𝑠
10
c. 𝐷𝑗 = { 𝑁+1
𝑗[ ]𝑡ℎ 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑑𝑖𝑠𝑐𝑟𝑒𝑡𝑒 𝑠𝑒𝑟𝑖𝑒𝑠.
10
j = 1, 2, … …9
𝑛+1 𝑡ℎ
𝑘[ ] 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙 𝑠𝑒𝑟𝑖𝑒𝑠
100
d. 𝑃𝑘 = { 𝑁+1
𝑘[ ]𝑡ℎ 𝑖𝑡𝑒𝑚 𝑓𝑜𝑟 𝑑𝑖𝑠𝑐𝑟𝑒𝑡𝑒 𝑠𝑒𝑟𝑖𝑒𝑠.
100
k = 1, 2, … … 99
For continuous series with exclusive type classes;
𝑁
a. Compute median index = and determine Md-class (the class corresponding
2
𝑁
𝑁 −𝑐𝑓
to the cf just greater than or equal to ). Then, 𝑀𝑑 = 𝑙 + 2 × ℎ where,
2 𝑓
l = lower limit of Md - class
cf = cumulative frequency preceding(अघिल्लो)the Md-class,
f = frequency of Md-class
h = size of Md-class
𝑖𝑁
b. Compute Qi index = and determine Qi -class (the class corresponding to the
4
𝑖𝑁
𝑖𝑁 − 𝑐𝑓
cf just greater than or equal to ). Then, 𝑄𝑖 = 𝑙 + 4
× ℎ where, 𝑖 = 1, 2, 3
4 𝑓
l = lower limit of Qi - class
cf = cumulative frequency preceding(अघिल्लो)the Qi - class,
f = frequency of Qi - class
h = size of Qi - class
𝑗𝑁
c. Compute Dj index = and determine Dj -class (the class corresponding to the
10
𝑗𝑁
𝑗𝑁 − 𝑐𝑓
cf just greater than or equal to ). Then, 𝐷𝑗 = 𝑙 + 10
×ℎ where, 𝑗 =
10 𝑓
1, 2, … … 9
l = lower limit of Dj - class
cf = cumulative frequency preceding(अघिल्लो)the Dj - class,
f = frequency of Dj - class
h = size of Dj - class
𝑘𝑁
d. Compute Pk index = and determine Pk -class (the class corresponding to
100
𝑘𝑁
𝑘𝑁 − 𝑐𝑓
the cf just greater than or equal to ). Then, 𝑃𝑘 = 𝑙 + 100 × ℎ where, 𝑘 =
100 𝑓
1, 2, … … 99
l = lower limit of Pk - class
cf = cumulative frequency preceding(अघिल्लो)the Pk - class,
f = frequency of Pk - class
h = size of Pk - class
Note: – The empirical relation between mean, median and mode for moderately asymmetrical
distribution is: 𝑀𝑜 = 3𝑀𝑑 − 2𝑥̅

You might also like