Standard Deviation

Standard deviation is the positive square root of the variance. It is one of the
basic methods of statistical analysis. Standard Deviation is commonly
abbreviated as SD and denoted by the symbol 'σ’ and it tells about how much
data values are deviated from the mean value. If we get a low standard
deviation then it means that the values tend to be close to the mean whereas
a high standard deviation tells us that the values are far from the mean value.
We have separate formulas to calculate the standard deviation of grouped
and ungrouped data. Also, we have different standard deviation formulas to
calculate SD of a random variable. Let us look into all the formulas in detail.
What is Standard Deviation?

Standard deviation is the degree of dispersion or the scatter of the data

points relative to its mean, in descriptive statistics. It tells how the values are
spread across the data sample and it is the measure of the variation of the
data points from the mean. The standard deviation of a data set, sample,
statistical population, random variable, or probability distribution is the square
root of its variance.
When we have n number of observations and the observations
are x1,x2,.....xn�1,�2,.....��, then the mean deviation of the value from the
mean is determined as ∑ni=1(xi−¯x)2∑�=1�(��−�¯)2. However, the sum of
squares of deviations from the mean doesn't seem to be a proper measure of
dispersion. If the average of the squared differences from the mean is small, it
indicates that the observations xi�� are close to the mean ¯x�¯. This is a
lower degree of dispersion. If this sum is large, it indicates that there is a
higher degree of dispersion of the observations from the mean ¯x�¯. Thus we
conclude that ∑ni=1(xi−¯x)2∑�=1�(��−�¯)2 is a reasonable indicator of the
degree of dispersion or scatter.
We take 1n∑ni=1(xi−¯x)21�∑�=1�(��−�¯)2 as a proper measure of
dispersion and this is called the variance(σ2). The positive square root of the
variance is the standard deviation.
Standard Deviation Formula

The spread of statistical data is measured by the standard deviation. The

degree of dispersion is computed by the method of estimating the deviation of
data points. You can read about dispersion in summary statistics. As discussed,
the variance of the data set is the average square distance between the mean
value and each data value. And standard deviation defines the spread of data
values around the mean. Here are two standard deviation formulas that are
used to find the standard deviation of sample data and the standard deviation
of the given population.
Note that both formulas look almost similar except for the denominator which
is N in the case of the population SD but n-1 in the case of the sample SD.
While calculating the sample mean, all the data values in the population are
not considered so the sample mean just is an estimate of the population
mean, but this introduces some uncertainty or bias in our calculation of
standard deviation. To adjust this, the denominator of the sample standard
deviation is corrected to be n-1 instead of just n. This is known as Bessel's
correction.
Formula for Calculating Standard Deviation

There are two types of data sets: populations and samples. A population is an
entire group that we are interested in studying, while a sample is a smaller
group of individuals that is taken from the population. The formulas to
calculate the standard deviations of population and sample differ a little.
The population standard deviation formula is given as:
σ=√1N∑Ni=1(Xi−μ)2�=1�∑�=1�(��−�)2
Here,
  σ = Population standard deviation symbol
 μ = Population mean
 N = total number of observations
Similarly, the sample standard deviation formula is:
s=√1n−1∑ni=1(xi−¯x)2�=1�−1∑�=1�(��−�¯)2
Here,
 s = Sample standard deviation symbol
 ¯x�¯ = Arithmetic mean of the observations
 n = total number of observations
How to Calculate Standard Deviation?

In general, the standard deviation refers to the population standard deviation

and here are the steps to calculate the standard deviation of a set of data
values:
 Find the mean, which is the arithmetic mean of the observations.
 Find the squared differences from the mean. (The data value - mean)2
 Find the average of the squared differences. (Variance = The sum of squared
differences ÷ the number of observations)
 Find the square root of variance. (Standard deviation = √Variance)
Standard Deviation of Ungrouped Data

The calculations for standard deviation differ for different data. Distribution measures the
deviation of data from its mean or average position. There are three methods to find the standard
deviation.

 Actual mean method

 Assumed mean method
 Step deviation method

Standard Deviation by The Actual Mean Method

In this method, we first compute the mean of the data values (¯x�¯) and then
compute the deviations of each data value from the mean. Then we use the
following standard deviation formula by actual mean method:
σ = √(∑(x−¯x)(�−�¯)2 /n), where n = total number of observations.
Consider the data observations 3, 2, 5, 6. Here the mean of these data points
is (3 + 2 + 5 + 6)/4 = 16/4 = 4.
The sum of the squared differences from mean = (4-3)2+(2-4)2 +(5-4)2 +(6-4)2 =
10
Variance = Squared differences from mean/ number of data points =10/4 =2.5
Standard deviation = √2.5 = 1.58
Standard deviation by Assumed Mean Method

When the x values are large, an arbitrary value (A) is chosen as the mean (as
the computation of mean is difficult in this case). The deviation from this
assumed mean is calculated as d = x - A. Then the standard deviation formula
by assumed mean method is:
σ = √[(∑(d)2 /n) - (∑d/n)2]
Standard Deviation by Step Deviation Method

The standard deviation of grouped data also can be calculated by "step

deviation method". In this method also, some arbitrary data value is chosen as
the assumed mean, A. Then we calculate the deviations of all data values by
using d = x - A. The next step is to calculate the step deviations (d') using d' =
d/i where 'i' is a common factor of all 'd' values (choose any common factor in
case of multiple factors). Now, the standard deviation of ungrouped data by
step deviation method is found by the formula:
σ = √[(∑(d')2 /n) - (∑d'/n)2] × i, where 'n' is the total number of data values.
Standard Deviation of Grouped Data (Discrete)
When the data points are grouped, we first construct a frequency
distribution. Just like ungrouped data, the standard deviation of grouped data
can also be calculated using 3 methods: actual mean method, assumed mean
method, and step deviation method. Let us dive into each of these methods.
Standard Deviation of Discrete Data by Actual Mean Method

For n number of observations, x1,x2,.....xn�1,�2,.....��, and the

corresponding frequencies, f1,f2,f3,...fn�1,�2,�3,...�� the standard deviation
is:
σ=√1n∑ni=1fi(xi−¯x)2�=1�∑�=1���(��−�¯)2. Here
 n = total frequency = ∑ni=1fi∑�=1���
 ¯x�¯ = mean
Example: Let's calculate the standard deviation for the data given below:

xi 6 10 12 14 24

fi 2 3 4 5 4

Calculate mean(¯x�¯): (6 × 2 + 10 × 3 + 12 × 4 + 14 × 5 + 24 ×
4)/(2+3+4+5+4) = 14.22
xi fi fixi xi - ¯x�¯ (xi - ¯x�¯)2 fi (xi - ¯x�¯)2

6 2 12 -8.22 67.5684 135.1368

10 3 24 -4.22 17.8084 53.4252

12 4 40 -2.22 4.9284 19.7136

14 5 60 -0.22 0.0484 0.242

24 4 56 9.78 95.6484 382.5936

  18 192     591.1112
Now, variance: σ2 = 1/n ∑ni=1fi(xi−¯x)2∑�=1���(��−�¯)2
= 1/18 × 591.1112 = 32.83
Calculate SD: σ = √Variance = √32.83 = 5.73
Standard Deviation of Discrete Data by Assumed Mean Method

When the data values are very large, then one of the data values is chosen as
the mean (and hence is known as assumed mean, A). Then the deviation of
each data value from the assumed mean is d = x - A. Then the formula for
standard deviation by assumed mean method is:
σ = √[(∑(fd)2 /n) - (∑fd/n)2], where
 'f' is the frequency of corresponding data value x and
 'n' is the total frequency.

Standard Deviation of Discrete Data by Step Deviation Method

The standard deviation of grouped data can be calculated by another method

called the "step deviation method". In this method also, we assume some data
value as the mean (assumed mean, A) and calculate the deviations of data
values using d = x - A. In this method, we also find the step deviations (d')
using d' = d/i where i = any number that is a common factor of all values
represented by 'd' (there can be multiple factors, but we can choose any).
Now, the standard deviation by step deviation method is calculated by the
formula:
σ = √[(∑(fd')2 /n) - (∑fd'/n)2] × i, where
 'f' is the frequency of data values
 'n' is the total frequency
 'i' is a common factor of all 'd' values, where d = x - A (A = assumed mean)
 d' = d/i
Standard Deviation of Grouped Data (Continuous)

If the frequency distribution is continuous, each class is replaced by its

midpoint. Then the standard deviation is calculated by the same technique as
in discrete frequency distribution. Consider the following example. xi�� is
calculated as the midpoint of each class which is calculated by the formula
(lower bound + upper bound)/2. Then the same standard deviation formula is
applied.
Class fi xi

0-10 3 5

10-20 4 15

20-30 6 25

30-40 4 35

40-50 8 40

Now, the standard deviation can be calculated by using the formulas of

grouped data either in the actual mean method, assumed mean method, or
step deviation method.
Standard Deviation of Random Variables

The measure of spread for the probability distribution of a random

variable determines the degree to which the values differ from the expected
value. This is a function that assigns a numerical value to each outcome in
a sample space. This is denoted by X, Y, or Z, as it is a function. If X is a
random variable, the standard deviation is determined by taking the square
root of the sum of the product of the squared difference between the random
variable, X, and the expected value (𝜇 (or) E(X)) and the probability associated
value of the random variable.
The standard deviation of the probability distribution of X, 𝜎
= √Σ[(x−μ)2⋅P(x)]Σ[(�−�)2⋅�(�)]
The shortcut to finding the standard deviation of random variables is: 𝜎
= √E(X2)−[E(X)]2�(�2)−[�(�)]2 (or) 𝜎 = √Σ[x2⋅P(x)]−μ2Σ[�2⋅�(�)]−�2
Standard Deviation of Probability Distribution

The experimental probability consists of many trials. When the difference

between the theoretical probability of an event and its relative frequency get
closer to each other, we tend to know the average outcome. This mean is
known as the expected value of the experiment denoted by 𝜇.
 In a normal distribution, the mean is zero and the standard deviation is 1.
 In a binomial experiment, the number of successes is a random variable. If a random
variable has a binomial distribution, its standard deviation is given by: 𝜎 = √npq,
where mean: 𝜇 = np, n = number of trials, p = probability of success and 1-p =q is the
probability of failure.
 In a Poisson distribution, the standard deviation is given by 𝜎= √λt, where λ is the
average number of successes in an interval of time t.

Important Notes on Standard Deviation:

 The square root of the average of the squared differences of data observations from
the mean is called the standard deviation.
 Standard deviation is the positive square root of variance.
 Standard deviation is the indicator that shows the dispersion of the data points about
the mean.