ARITHMETIC MEAN:

∑ 𝑥𝑖
֎ Ungrouped Data (Direct) : 𝑋= 𝑛
Where, xi = each observation
n = total number of observations

∑ 𝑑𝑖
֎ Ungrouped Data (Shortcut) : 𝑋=𝐴+ 𝑛
Where, A = assumed mean
di = xi – A
n = total number of observations

∑ 𝑓 𝑖 𝑋𝑖
֎ Discrete Data (Direct) : 𝑋= 𝑛
Where, fi = frequency of each observation
n = total number of observations

∑ 𝑓𝑖 𝑑𝑖
֎ Discrete Data (Shortcut) : 𝑋=𝐴+ 𝑛
Where, A = assumed mean
fi = frequency of each observation
di = xi – A
n = total number of observations

∑ 𝑓𝑖 𝑚𝑖
֎ Continuous Data (Direct) : 𝑋= 𝑛
Where, mi = middle value
n = total number of observations

∑ 𝑓𝑖 𝑑𝑖
֎ Continuous Data (Step Deviation) : 𝑋=𝐴+ ( )∗ℎ
𝑛
Where, A = assumed mean
𝑚𝑖 −𝐴
𝑑𝑖 =
ℎ
h = width of the class
n = total number of observations

∑ 𝑋𝑖 𝑤𝑖
֎ Weighted Mean : 𝑋𝑤 = ∑ 𝑤𝑖
Where, wi = weight of the observations

∑ 𝑛𝑖 𝑋𝑖
֎ Combined Arithmetic Mean : 𝑋𝑐 = ∑ 𝑛𝑖
Where, 𝑋𝑖 = Individual arithmetic mean
ni = number of observations in each A.M
GEOMETRIC MEAN:
∑ 𝑙𝑜𝑔𝑋𝑖
֎ Ungrouped Data : 𝐺𝑀 = 𝐴𝑛𝑡𝑖𝑙𝑜𝑔 ( )
𝑛
Where, Xi = each observation
n = total number of observations

∑ 𝑓𝑖 𝑙𝑜𝑔 𝑋𝑖
֎ Discrete Data : 𝐺𝑀 = 𝑛
Where, fi = frequency of each observation
n = total number of observations

∑ 𝑓𝑖 𝑙𝑜𝑔 𝑚𝑖
֎ Continuous Data : 𝐺𝑀 = 𝑛
Where, mi = middle value
n = total number of observations

∑ 𝑙𝑜𝑔𝑋𝑖 𝑤𝑖
֎ Weighted Geometric Mean : 𝐺𝑀𝑤 = ∑ 𝑤𝑖
Where, wi = weight of the observations

∑ 𝑛𝑖 𝑙𝑜𝑔𝐺𝑖
֎ Combined Geometric Mean : 𝐺𝑀 = ∑ 𝑛𝑖
Where, Gi = individual GM
ni = number of observations in each GM

HARMONIC MEAN:
𝑁
֎ Ungrouped Data : 𝐻𝑀 = 1
∑( )
𝑥𝑖

Where, xi = each observation

N = total number of observations

𝑁
֎ Discrete Data : HM = 𝑓
∑( 𝑖 )
𝑥𝑖

Where, fi = frequency of each observation

xi = each observation
N = total number of observations

𝑁
֎ Continuous Data : HM = 𝑓
∑( 𝑖 )
𝑚𝑖

Where, fi = frequency of each observation

mi = middle value
n = total number of observations
∑ 𝑊𝑖
֎ Weighted Harmonic Mean : 𝐻𝑀𝑤 = 𝑤
∑( 𝑖 )
𝑥𝑖

Where, wi = weight of the observations

xi = each observation
MEDIAN:
𝑁+1
֎ Individual Observations (if N is odd) : 𝑀𝑒𝑑 = 𝑆𝑖𝑧𝑒 𝑜𝑓 ( ) 𝑡ℎ 𝑖𝑡𝑒𝑚
2
Where, N = total number of observations

OR

𝑁 𝑁+1
+
֎ Individual Observations (if N is even) : 𝑀𝑒𝑑 = 𝑆𝑖𝑧𝑒 𝑜𝑓 ( 2 2
) 𝑡ℎ 𝑖𝑡𝑒𝑚
2

Where, N = total number of observations

(𝑁⁄2)−𝑐.𝑓
֎ Continuous Data : 𝑀𝑒𝑑 = 𝐿 + [ ]∗𝑖
𝑓
Where, L = lower limit of median class
c.f = cumulative frequency upto but not including the median class
i = width of median class
N = total number of observations

MODE:
∆1
֎ Continuous Data : 𝑀𝑜𝑑𝑒 = 𝐿 + [∆ ]∗𝑖
1 + ∆2
Where, L = lower limit of modal class
Δ1 = difference between frequency of modal class and pre-modal class
Δ2 = difference between frequency of modal class and post-modal class
i = width of modal class

RELATIONSHIP BETWEEN MEAN, MEDIAN, AND MODE:

𝑀𝑜𝑑𝑒 = (3 ∗ 𝑀𝑒𝑑𝑖𝑎𝑛) − (2 ∗ 𝑀𝑒𝑎𝑛)

QUARTILES:
𝑁+1
֎ Ungrouped Data (First Quartile) : 𝑄1 = 𝑆𝑖𝑧𝑒 𝑜𝑓 ( ) 𝑡ℎ 𝑖𝑡𝑒𝑚
4

(𝑁⁄ )−𝑐.𝑓
֎ Grouped Data (First Quartile) : 𝑄1 = 𝐿+ [ 4 ]∗𝑖
𝑓

𝑁+1
֎ Ungrouped Data (Third Quartile) : 𝑄1 = 𝑆𝑖𝑧𝑒 𝑜𝑓 3 ∗ ( ) 𝑡ℎ 𝑖𝑡𝑒𝑚
4

(3𝑁⁄ )−𝑐.𝑓
֎ Grouped Data (Third Quartile) : 𝑄1 = 𝐿+ [ 4 ]∗𝑖
𝑓
DECILES:
𝑁+1
֎ Ungrouped Data : 𝐷𝑗 = 𝑆𝑖𝑧𝑒 𝑜𝑓 𝑗 ∗ ( ) 𝑡ℎ 𝑖𝑡𝑒𝑚
10

(
𝑗∗𝑁⁄ )−𝑐.𝑓
10
֎ Grouped Data : 𝐷𝑗 = 𝐿+ [ 𝑓
]∗𝑖

Where j = Decile required

L = lower limit of decile class
c.f = cumulative frequency upto but not including the median class
i = width of median class
N = total number of observations

PERCENTILES:
𝑁+1
֎ Ungrouped Data : 𝐷𝑗 = 𝑆𝑖𝑧𝑒 𝑜𝑓 𝑗 ∗ ( ) 𝑡ℎ 𝑖𝑡𝑒𝑚
10

(
𝑗∗𝑁⁄ )−𝑐.𝑓
100
֎ Grouped Data : 𝐷𝑗 = 𝐿+ [ 𝑓
]∗𝑖

Where j = Percentile required

L = lower limit of percentile class
c.f = cumulative frequency upto but not including the median class
i = width of median class
N = total number of observations
 Measures of Dispersion
RANGE:
𝑅𝑎𝑛𝑔𝑒 = 𝐿 − 𝑆

L = largest item
S = smallest item

COEFFICIENT OF RANGE:
𝐿−𝑆
𝐶𝑜 − 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡𝑅𝑎𝑛𝑔𝑒 =
𝐿+𝑆

INTERQUARTILE RANGE:
𝐼𝑛𝑡𝑒𝑟𝑞𝑢𝑎𝑟𝑖𝑙𝑡𝑒 𝑅𝑎𝑛𝑔𝑒 = 𝑄3 − 𝑄1

QUARTILE DEVIATION:
𝑄3 − 𝑄1
𝑄. 𝐷. =
2

COEFFICIENT OF QUARTILE DEVIATION:

𝑄3 − 𝑄1
𝐶𝑜 − 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡𝑄.𝐷. =
𝑄3 + 𝑄1

MEAN DEVIATION:
∑|𝐷|
֎ Ungrouped Data : 𝑀. 𝐷. =
𝑁

Where, |D| = Absolute value of deviations from mean

N = No of observations
∑ 𝑓|𝐷|
֎ Discrete/Continuous Series : 𝑀. 𝐷. =
𝑁

Where, f = frequency
COEFFICIENT OF MEAN DEVIATION:

𝑀. 𝐷.
𝐶𝑜 − 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡𝑀.𝐷. =
𝑀𝑒𝑑𝑖𝑎𝑛

OR

𝑀. 𝐷.
𝐶𝑜 − 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡𝑀.𝐷. =
𝑀𝑒𝑎𝑛

STANDARD DEVIATION:
∑ 𝑥2
֎ Ungrouped Data (Actual Mean) : 𝜎= √
𝑁

Where, 𝑥 = (𝑋 − 𝑋̅ )

N = No of observations

∑𝑑 ∑𝑑2 2
֎ Ungrouped Data (Assumed Mean) : 𝜎= √ − ( )
𝑁 𝑁

Where, 𝑑 = (𝑋 − 𝐴)

∑ 𝑓𝑥 2
֎ Discrete Series (Actual Mean) : 𝜎= √
𝑁

Where, 𝑥 = (𝑋 − 𝑋̅ )

N = No of observations

∑ 𝑓𝑑 2
∑ 𝑓𝑑 2
֎ Discrete Series (Assumed Mean) : 𝜎= √ − ( )
𝑁 𝑁

Where, 𝑑 = (𝑋 − 𝐴)

∑ 𝑓𝑑 ′2
∑ 𝑓𝑑 ′ 2
֎ Discrete Series (Step Deviation) : 𝜎= √ − ( ) ∗𝐶
𝑁 𝑁

(𝑋− 𝐴)
Where, 𝑑 ′ =
𝐶

C = Common factor

∑ 𝑓𝑥 2
֎ Continuous Series (Actual Mean) : 𝜎= √
𝑁

Where, 𝑥 = (𝑚 − 𝑋̅)

N = No of observations

∑ 𝑓𝑑 2
∑ 𝑓𝑑 2
֎ Continuous Series (Assumed Mean) : 𝜎= √ − ( )
𝑁 𝑁

Where, 𝑑 = (𝑚 − 𝐴)
 ∑ 𝑓𝑑 ′2∑ 𝑓𝑑 ′ 2
֎ Continuous Series (Step Deviation) : 𝜎= √ − ( ) ∗𝐶
𝑁 𝑁

(𝑚− 𝐴)
Where, 𝑑 ′ =
𝐶

C = Common factor

COMBINED STANDARD DEVIATION:

𝑁1 𝜎12 + 𝑁2 𝜎22 + 𝑁1 𝑑12 + 𝑁2 𝑑22
𝜎12 = √
𝑁1 + 𝑁2

Where, 𝑑1 = (𝑋̅1 − 𝑋̅12 )

𝑑2 = (𝑋̅2 − 𝑋̅12 )

N = No of observations

COEFFICIENT OF VARIATION:

𝜎
𝐶. 𝑉 = ( ) ∗ 100
𝑋̅

VARIANCE:

∑(𝑋− 𝑋̅)2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 = = 𝜎2
𝑁

OR

𝜎 = √𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒
 Skewness, Moments & Kurtosis
SKEWNESS (KARL PEARSON’S COEFFICIENT):

𝑋̅ − 𝑀𝑜𝑑𝑒
𝑆𝑘 =
𝜎
Where, 𝑋̅ = Arithmetic Mean
S = smallest item

OR

3 ∗ (𝑋̅ − 𝑀𝑒𝑑𝑖𝑎𝑛)
𝑆𝑘 =
𝜎

SKEWNESS (BOWLEY’S COEFFICIENT):

(𝑄3 + 𝑄1 ) − 2 ∗ 𝑀𝑒𝑑𝑖𝑎𝑛
𝑆𝑘 =
𝑄3 − 𝑄1

Where, 𝑄1 = First Quartile

𝑄3 = Third Quartile

SKEWNESS (KELLY’S COEFFICIENT):

(𝑃10 + 𝑃90 ) − 2 ∗ 𝑀𝑒𝑑𝑖𝑎𝑛

𝑆𝑘 =
𝑃90 − 𝑃10

Where, 𝑃10 = 10th Percentile

𝑃90 = 90th Percentile

OR

(𝐷1 + 𝐷9 ) − 2 ∗ 𝑀𝑒𝑑𝑖𝑎𝑛
𝑆𝑘 =
𝐷9 − 𝐷1

Where, 𝐷1 = 1st Decile

𝐷9 = 9th Decile
MOMENTS

֎ Moments are the average of each power of deviations from the arithmetic mean of a distribution.
֎ The first four moments are generally considered for statistical inferences.
֎ The first moment describes arithmetic mean.
֎ The second moment describes variance and standard deviation.
֎ The third moment describes skewness.
֎ The fourth moment describes kurtosis.

Ungrouped Data Discrete/Continuous Series

∑(𝑋 − 𝑋̅) ∑ 𝑓(𝑋 − 𝑋̅)
𝜇1 = 𝜇1 =
𝑁 𝑁
2 2
∑(𝑋 − 𝑋̅) ∑ 𝑓(𝑋 − 𝑋̅)
𝜇2 = 𝜇2 =
𝑁 𝑁
3 3
∑(𝑋 − 𝑋̅) ∑ 𝑓(𝑋 − 𝑋̅)
𝜇3 = 𝜇3 =
𝑁 𝑁
4 4
∑(𝑋 − 𝑋̅) ∑ 𝑓(𝑋 − 𝑋̅)
𝜇4 = 𝜇4 =
𝑁 𝑁

Where, 𝑋̅ = Arithmetic Mean

N = Number of observations
𝜇32
֎ Skewness : 𝛽1 =
𝜇23

֎ If data is normally distributed, 𝛽1 = 0

֎ If data is positively skewed, 𝛽1 > 0
֎ If data is negatively skewed, 𝛽1 < 0

KURTOSIS:
𝜇4
֎ Kurtosis : 𝛽2 =
𝜇22

֎ If data is mesokurtic, 𝛽2 = 3
֎ If data is leptokurtic, 𝛽2 > 3
֎ If data is platykurtic, 𝛽2 < 3
 Index Numbers
UNWEIGHTED PRICE INDICES:

Simple Aggregative Price Index:

∑ 𝑝1
𝑃01 = ∗ 100
∑ 𝑝0

֎ P01 = Index number of current year

֎ P1 = Total of current year prices for all commodities
֎ P0 = Total of base year prices for all commodities

Simple Average of Price Relatives:

𝑝
∑ ( 1 ) ∗ 100
𝑝0
𝑃01 =
𝑁
𝑝
∑ 𝑙𝑜𝑔 {( 1 ) ∗ 100}
𝑝0
𝑙𝑜𝑔 𝑃01 =
𝑁
𝑝
(∑ 𝑙𝑜𝑔 ( 1 ) ∗ 100) ∑ 𝑙𝑜𝑔 𝑃
𝑝0
𝑃01 = 𝑎𝑛𝑡𝑖𝑙𝑜𝑔 [ ] = 𝑎𝑛𝑡𝑖𝑙𝑜𝑔
𝑁 𝑁

֎ P01 = Index number of current year

֎ P1 = Total of current year prices for all commodities
֎ P0 = Total of base year prices for all commodities

WEIGHTED PRICE INDICES:

Laspeyres Method:

∑ 𝑝1 𝑞0
𝑃01 = ∗ 100
∑ 𝑝0 𝑞0
Paasche’s Method:

∑ 𝑝1 𝑞1
𝑃01 = ∗ 100
∑ 𝑝0 𝑞1

Dorbish & Bowley’s Method:

1 ∑ 𝑝1 𝑞0 ∑ 𝑝1 𝑞1
𝑃01 = { + } ∗ 100
2 ∑ 𝑝0 𝑞0 ∑ 𝑝0 𝑞1

Fisher’s Ideal Method:

1 ∑ 𝑝1 𝑞0 ∑ 𝑝1 𝑞1
𝑃01 = { + } ∗ 100
2 ∑ 𝑝0 𝑞0 ∑ 𝑝0 𝑞1

Marshall-Edgeworth Method:

∑(𝑞0 + 𝑞1 ) ∗ 𝑝1 ∑ 𝑞0 𝑝1 + ∑ 𝑞1 𝑝1
𝑃01 = ∗ 100 = ∗ 100
∑(𝑞0 + 𝑞1 ) ∗ 𝑝0 ∑ 𝑞0 𝑝0 + ∑ 𝑞1 𝑝0

Kelly’s Method:

∑ 𝑝1 𝑞
𝑃01 = ∗ 100
∑ 𝑝0 𝑞

Where q = fixed weight

𝑞0 + 𝑞1
𝑞=
2

֎ p1 = Total of current year prices for all commodities

֎ p0 = Total of base year prices for all commodities
֎ q1 = Total of current year quantities for all commodities
֎ q0 = Total of base year quantities for all commodities