Formulas for Measures of Central Tendency
Statistical Methods
Introduction
This document presents the formulas for various measures of central tendency including
arithmetic mean, geometric mean, harmonic mean, median, and mode for both grouped
and ungrouped data.
1 Arithmetic Mean
1.1 Ungrouped Data
1.1.1 Raw Data
For n observations x1 , x2 , . . . , xn : Pn
i=1 xi
x̄ =
n
1.1.2 Discrete Frequency Distribution
For values x1 , x2 , . . . , xk with frequencies f1 , f2 , . . . , fk :
Pk P
f i x i f i xi
x̄ = Pi=1
k
=
i=1 fi
N
where N = ki=1 fi is the total frequency.
P
1.2 Grouped Data
For grouped data with k classes having midpoints m1 , m2 , . . . , mk and frequencies f1 , f2 , . . . , fk :
Pk P
i=1 fi mi fi mi
x̄ = Pk =
i=1 fi
N
2 Geometric Mean
2.1 Ungrouped Data
2.1.1 Raw Data
For n positive observations x1 , x2 , . . . , xn :
n
! n1
Y
GM = xi
i=1
1
Measures of Central Tendency 2
Alternatively: !
n
1X
GM = exp ln xi
n i=1
2.1.2 Discrete Frequency Distribution
For positive values x1 , x2 , . . . , xk with frequencies f1 , f2 , . . . , fk :
k
! Pk 1 k
! N1
f
Y i=1 i Y
GM = xfi i = xfi i
i=1 i=1
2.2 Grouped Data
For grouped data with k classes having midpoints m1 , m2 , . . . , mk and frequencies f1 , f2 , . . . , fk :
k
! Pk 1 k
! N1
f
Y i=1 i Y
GM = mfi i = mfi i
i=1 i=1
3 Harmonic Mean
3.1 Ungrouped Data
3.1.1 Raw Data
For n non-zero observations x1 , x2 , . . . , xn :
n
HM = Pn 1
i=1 xi
3.1.2 Discrete Frequency Distribution
For non-zero values x1 , x2 , . . . , xk with frequencies f1 , f2 , . . . , fk :
Pk
fi N
HM = Pki=1 f = Pk f
i i
i=1 xi i=1 xi
3.2 Grouped Data
For grouped data with k classes having midpoints m1 , m2 , . . . , mk and frequencies f1 , f2 , . . . , fk :
Pk
fi N
HM = Pki=1 f = Pk f
i i
i=1 mi i=1 mi
Measures of Central Tendency 3
4 Median
4.1 Ungrouped Data
4.1.1 Raw Data
Let x(1) ≤ x(2) ≤ · · · ≤ x(n) be the ordered observations:
x n+1 if n is odd
( )
Median = 1 h 2 i
x n +x n
2 ( )
2 ( +1) if n is even
2
4.1.2 Discrete Frequency Distribution
For cumulative frequency distribution:
N +1
Median = Value corresponding to the -th observation
2
Pk
where N = i=1 fi .
4.2 Grouped Data
For grouped data with continuous classes:
!
N
2
−F
Median = L + ×c
f
where:
• L = lower boundary of median class
• N = total frequency ( fi )
P
• F = cumulative frequency of the class preceding the median class
• f = frequency of the median class
• c = class width
5 Mode
5.1 Ungrouped Data
5.1.1 Raw Data
Mode = Value that occurs with maximum frequency
5.1.2 Discrete Frequency Distribution
Mode = Value corresponding to the maximum frequency
Measures of Central Tendency 4
5.2 Grouped Data
For grouped data with continuous classes:
f1 − f 0
Mode = L + ×c
2f1 − f0 − f2
where:
• L = lower boundary of modal class
• f1 = frequency of modal class
• f0 = frequency of class preceding modal class
• f2 = frequency of class succeeding modal class
• c = class width
5.3 Empirical Relationship
For moderately skewed distributions:
Mode = 3 × Median − 2 × Mean
Important Notes
1. The geometric mean and harmonic mean are defined only for positive values.
2. For grouped data formulas, all classes are assumed to be of equal width.
3. n denotes the number of observations in ungrouped data.
P
4. N = fi denotes the total frequency in grouped/frequency data.
N
5. The median class is the class containing the 2
-th observation.
6. The modal class is the class with the highest frequency.
7. The relationship among the three measures of central tendency for symmetrical
distributions is:
Mean = Median = Mode
8. For moderately skewed distributions, the empirical relationship is approximately
valid.
Measures of Central Tendency 5
Symbols Used
Symbol Meaning
xi i-th observation
fi frequency of i-th value/class
mi midpoint of i-th class
n number of observations
P (ungrouped data)
N total frequency ( fi )
k number of distinct values/classes
x̄ arithmetic mean
GM geometric mean
HM
Q harmonic mean
P product symbol
summation symbol
exp(x) ex
ln(x) natural logarithm of x
End of Document