n

x i
Mean x  i 1

n
n

fx i 1
i i
Mean x 
f i

  f i ui  x A
Step Deviation Method: Mean = A  h  where u i  i
 f  h
 i 

fd i i
Assumed Mean Method: Mean 𝑥̅ = 𝐴 + i 1
where di   x  A
f i

th th
n n 
  item    1 item
 n 1
th

Median =  2 2 
 item if n is odd or if n is even
 2  2

𝒏
− 𝒄𝒇
𝟐
Median = 𝒍+ ( )×𝒉
𝒇

[cf = cumulative frequency of previous class, frequency of median class(n/2) th class]

Quartiles are the values which divide the data in 4 equal parts
Ungrouped data
th th
n n 
  item    1 item
𝑸𝟏 =
𝒏+𝟏
𝒕𝒉 𝒐𝒃𝒔𝒆𝒓𝒗𝒂𝒕𝒊𝒐𝒏 n is odd 4 4  if n is even
𝟒 2

th th
n n 
  item    1 item
𝒕𝒉 𝒐𝒃𝒔𝒆𝒓𝒗𝒂𝒕𝒊𝒐𝒏   2 
𝟐(𝒏+𝟏) 2
𝑸𝟐 = if n is even
𝟒 2
 th th
 3n   3n 
  item    1 item
𝒕𝒉 𝒐𝒃𝒔𝒆𝒓𝒗𝒂𝒕𝒊𝒐𝒏    4 
𝟑(𝒏+𝟏) 4
𝑸𝟑 = if n is even
𝟒 2

𝑸𝟑 −𝑸𝟏
Quartile Deviation =
𝟐

Grouped Data

𝒏
− 𝒄𝒇
𝟒
𝑸𝟏 = 𝒍 + ( )×𝒉
𝒇

𝒏
− 𝒄𝒇
𝟐
𝑸𝟐 = 𝒍 + ( ) × 𝒉 (Median)
𝒇

𝟑𝒏
− 𝒄𝒇
𝟒
𝑸𝟑 = 𝒍 + ( )×𝒉
𝒇

𝒇𝟏 −𝒇𝟎
Mode = 𝒍 + ( )×𝒉
𝟐𝒇𝟏 −𝒇𝟎 −𝒇𝟐

𝒇𝟏 = 𝒇𝒓𝒆𝒒𝒖𝒆𝒏𝒄𝒚 𝒐𝒇 𝒎𝒐𝒅𝒂𝒍 𝒄𝒍𝒂𝒔𝒔, 𝒇𝟎 = 𝒇𝒓𝒆𝒒𝒖𝒆𝒏𝒄𝒚 𝒐𝒇 𝒄𝒍𝒂𝒔𝒔 𝒑𝒓𝒆𝒄𝒆𝒆𝒅𝒊𝒏𝒈 𝒎𝒐𝒅𝒂𝒍 𝒄𝒍𝒂𝒔𝒔, 𝒇𝟐 =

𝒇𝒓𝒆𝒒𝒖𝒆𝒏𝒄𝒚 𝒐𝒇 𝒄𝒍𝒂𝒔𝒔 𝒔𝒖𝒄𝒄𝒆𝒆𝒅𝒊𝒏𝒈 𝒕𝒉𝒆 𝒎𝒐𝒅𝒂𝒍 𝒄𝒍𝒂𝒔𝒔

Mode = 3 Median – 2 Mean

Range = Maximum Value – Minimum Value

∑𝒏
𝒊=𝟏|𝒙𝒊 −𝒙
̅|
̅)
Mean Deviation about Mean: MD (𝒙 =
𝒏
∑𝒏
𝒊=𝟏 𝒇𝒊 |𝒙𝒊 −𝒙
̅|
̅) =
MD (𝒙
∑ 𝒇𝒊
 ∑𝒏
𝒊=𝟏|𝒙𝒊 −𝑴|
Mean Deviation about Median: MD (𝐌) =
𝒏
∑𝒏
𝒊=𝟏 𝒇𝒊 |𝒙𝒊 −𝑴|
MD (M) =
∑ 𝒇𝒊

Standard Deviation 𝝈 = √𝑽𝒂𝒓𝒊𝒂𝒏𝒄𝒆

𝟏 𝟏
𝝈 = √ ∑𝒏𝒊=𝟏(𝒙𝒊 − 𝒙
̅ )𝟐 or 𝝈 = √ ∑ 𝒙𝒊 𝟐 − 𝒙
̅𝟐
𝒏 𝒏

𝟏 ∑𝒅 𝟐
𝝈 = √ ∑ 𝒅𝒊 𝟐 − ( 𝒊 )
𝒏 𝒏

𝟏
𝝈 = √ ∑𝒏𝒊=𝟏 𝒇𝒊 (𝒙𝒊 − 𝒙
̅ )𝟐
𝑵

𝟏
𝝈=
𝑵
√𝑵 ∑𝒏𝒊=𝟏 𝒇𝒊 𝒙𝒊 𝟐 − (∑𝒏𝒊=𝟏 𝒇𝒊 𝒙𝒊 )𝟐

𝒉 𝟐 xi  A
𝝈= 𝑵
√𝑵 ∑𝒏
𝒊=𝟏 𝒇𝒊 𝒖𝒊 𝟐 − (∑𝒏𝒊=𝟏 𝒇𝒊 𝒖𝒊) where u i 
h
𝟐
∑ 𝒇𝒊 𝒖𝒊 𝟐 ∑ 𝒇𝒊 𝒖𝒊
𝝈 = 𝒉√ − ( )
∑ 𝒇𝒊 ∑ 𝒇𝒊

𝟐 𝒏 𝟐
∑ 𝒇𝒊 𝒅𝒊 ∑ 𝒇𝒅
𝝈= √ 𝑵 − ( 𝒊=𝟏𝑵 𝒊 𝒊 ) where 𝒅𝒊 = 𝒙𝒊 − 𝑨

Skewness and Kurtosis

Karl Pearson Coefficient of Skewness 𝑺𝑲 𝑴𝒆𝒂𝒏−𝑴𝒐𝒅𝒆 𝟑(𝑴𝒆𝒂𝒏−𝑴𝒆𝒅𝒊𝒂𝒏)

𝒑 = 𝑺𝒕𝒂𝒏𝒅𝒂𝒓𝒅 𝑫𝒆𝒗𝒊𝒂𝒕𝒊𝒐𝒏 𝑶𝑹
𝑺𝒕𝒂𝒏𝒅𝒂𝒓𝒅 𝑫𝒆𝒗𝒊𝒂𝒕𝒊𝒐𝒏

Bowley’s coefficient of Skewness 𝑺 𝑸 +𝑸 −𝟐𝑸𝟐

𝑲𝑩 = 𝟑 𝟏
𝑸𝟑 −𝑸𝟏

𝑳 + 𝟎.𝟓 𝑬
Percentile Rank for ungrouped data: 𝑷𝑹 = × 𝟏𝟎𝟎
𝒏
 𝑪 + 𝟎.𝟓 𝒇
Percentile Rank for grouped data: 𝑷𝑹 = ∑𝒇
× 𝟏𝟎𝟎

∑(𝒙−𝒙 ̅)
̅)(𝒚−𝒚
Cov (X,Y) or 𝑪𝑿𝒀 = 𝑵

𝟏 𝟏
Cov (X,Y) = 𝑵 [∑ 𝒙𝒚 − 𝑵 ∑ 𝒙 ∑ 𝒚]

𝟏 𝟏
Cov( X, Y) = [∑ 𝒖𝒗 − 𝑵 ∑ 𝒖 ∑ 𝒗] 𝒖 = 𝒙 − 𝑨, 𝒗 = 𝒚 − 𝑩
𝑵

Karl Pearson’s Coefficient of Correlation

𝑪𝒐𝒗(𝒙, 𝒚)
𝒓=
𝝈𝒙 𝝈𝒚

∑(𝒙 − 𝒙
̅)(𝒚 − 𝒚
̅)
𝒓=
̅)𝟐 √∑(𝒚 − 𝒚
√∑(𝒙 − 𝒙 ̅ )𝟐

𝟏
∑ 𝒙𝒚 − ∑ ∑
𝒓= 𝑵 𝒙 𝒚
√∑ 𝒙𝟐 − 𝟏 (∑ 𝒙)𝟐 √∑ 𝒚𝟐 − 𝟏 (∑ 𝒚)𝟐
𝑵 𝑵

𝟏
∑ 𝒖𝒗 − ∑ ∑
𝒓= 𝑵 𝒖 𝒗
√∑ 𝒖𝟐 − 𝟏 (∑ 𝒖)𝟐 √∑ 𝒗𝟐 − 𝟏 (∑ 𝒗)𝟐
𝑵 𝑵

Spearman’s rank correlation coefficient

𝟏
𝟔 ∑ 𝒅𝟐 𝟔[∑ 𝒅𝟐 + (𝒎𝟑 −𝒎)]
𝟏𝟐
𝒓=𝟏− OR 𝒓 =𝟏−
𝒏(𝒏𝟐 −𝟏) 𝒏(𝒏𝟐 −𝟏)