Statistics is a branch of mathematics that deals with

collecting, interpreting, organization and interpretation

of data.

Initially, when we get the data, instead of applying fancy

algorithms and making some predictions, we first try to read and
understand the data by applying statistical techniques. By doing
this, we are able to understand what type of distribution data has.

This blog aims to answer following questions:

1. What is Descriptive Statistics?

2. Types of Descriptive Statistics?

3. Measure of Central Tendency (Mean, Median, Mode)

4. Measure of Spread / Dispersion (Standard Deviation, Mean

Deviation, Variance, Percentile, Quartiles, Interquartile Range)

5. What is Skewness?

6. What is Kurtosis?

7. What is Correlation?

Today, let’s understand descriptive statistics once and for all. Let’s
start,
What is Descriptive Statistics?
Descriptive statistics involves summarizing and organizing the
data so they can be easily understood. Descriptive statistics, unlike
inferential statistics, seeks to describe the data, but do not attempt
to make inferences from the sample to the whole population. Here,
we typically describe the data in a sample. This generally means
that descriptive statistics, unlike inferential statistics, is not
developed on the basis of probability theory.

Types of Descriptive Statistics?

Descriptive statistics are broken down into two categories.
Measures of central tendency and measures of variability (spread).

Measure of Central Tendency

Central tendency refers to the idea that there is one number that
best summarizes the entire set of measurements, a number that is
in some way “central” to the set.

Mean / Average
Mean or Average is a central tendency of the data i.e. a number
around which a whole data is spread out. In a way, it is a single
number which can estimate the value of whole data set.

Let’s calculate mean of the data set having 8 integers.

Image 2

Median
Median is the value which divides the data in 2 equal parts i.e.
number of terms on right side of it is same as number of terms on
left side of it when data is arranged in either ascending or
descending order.

Note: If you sort data in descending order, it won’t affect median

but IQR will be negative. We will talk about IQR later in this blog.

Median will be a middle term, if number of terms is odd

Median will be average of middle 2 terms, if number of terms is

even.

Image 3

The median is 59 which will divide set of numbers into equal two
parts. Since there are even numbers in the set, the answer is
average of middle numbers 51 and 67.

Note: When values are in arithmetic progression (difference

between the consecutive terms is constant. Here it is 2.), median
is always equal to mean.

Image 4

An mean of these 5 numbers is 6 and so median.

Mode
Mode is the term appearing maximum time in data set i.e. term
that has highest frequency.

Image 5

In this data set, mode is 67 because it has more than rest of the
values, i.e. twice.

But there could be a data set where there is no mode at all as all
values appears same number of times. If two values appeared
same time and more than the rest of the values then the data set
is bimodal. If three values appeared same time and more than
the rest of the values then the data set is trimodal and for n
modes, that data set is multimodal.

Measure of Spread / Dispersion

Measure of Spread refers to the idea of variability within your
data.

Standard deviation
Standard deviation is the measurement of average distance
between each quantity and mean. That is, how data is spread out
from mean. A low standard deviation indicates that the data points
tend to be close to the mean of the data set, while a high standard
deviation indicates that the data points are spread out over a wider
range of values.

There are situations when we have to choose between sample or

population Standard Deviation.
When we are asked to find SD of some part of a population, a
segment of population; then we use sample Standard Deviation.

Image 6

where x̅ is mean of a sample.

But when we have to deal with a whole population, then we use

population Standard Deviation.

Image 7

where µ is mean of a population.

Though sample is a part of a population, their SD formulas should

have been same, but it is not. To find out more about it, refer
this link

As you know, in descriptive statistics, we generally deal with a data

available in a sample, not in a population. So if we use previous
data set, and substitute the values in sample formula,

Image 8
And answer is 29.62.

Mean Deviation / Mean Absolute Deviation

It is an average of absolute differences between each value in a set
of values, and the average of all values of that set.

Mean Deviation [Image 9] (Image courtesy: My Photoshopped Collection)

So if we use previous data set, and substitute the values,

Image 10

And answer is 23.75.

Variance
Variance is a square of average distance between each quantity
and mean. That is it is square of standard deviation.

Image 11

And answer is 877.34.

Range
Range is one of the simplest techniques of descriptive statistics. It
is the difference between lowest and highest value.

Image 12

Range is 99–12 = 87

Percentile
Percentile is a way to represent position of a values in data set. To
calculate percentile, values in data set should always be in
ascending order.

Image 13

The median 59 has 4 values less than itself out of 8. It can also be
said as: In data set, 59 is 50th percentile because 50% of the total
terms are less than 59. In general, if k is nth percentile, it implies
that n% of the total terms are less than k.

Quartiles
In statistics and probability, quartiles are values that divide your
data into quarters provided data is sorted in an ascending
order.

Quartiles [Image 14] (Image courtesy: https://statsmethods.wordpress.com/2013/05/09/iqr/)

There are three quartile values. First quartile value is at 25
percentile. Second quartile is 50 percentile and third quartile is 75
percentile. Second quartile (Q2) is median of the whole data. First
quartile (Q1) is median of upper half of the data. And Third
Quartile (Q3) is median of lower half of the data.

So here, by analogy,

Q2 = 67: is 50 percentile of the whole data and is median.

Q1 = 41: is 25 percentile of the data.

Q3 = 85: is 75 percentile of the date.

Interquartile range (IQR) = Q3 - Q1 = 85 - 41 = 44

Note: If you sort data in descending order, IQR will be -44. The
magnitude will be same, just sign will differ. Negative IQR is fine,
if your data is in descending order. It just we negate smaller values
from larger values, we prefer ascending order (Q3 - Q1).

Skewness
Skewness is a measure of the asymmetry of the probability
distribution of a real-valued random variable about its mean. The
skewness value can be positive or negative, or undefined.

In a perfect normal distribution, the tails on either side of the

curve are exact mirror images of each other.
When a distribution is skewed to the left, the tail on the curve’s
left-hand side is longer than the tail on the right-hand side, and
the mean is less than the mode. This situation is also called
negative skewness.

When a distribution is skewed to the right, the tail on the curve’s

right-hand side is longer than the tail on the left-hand side, and
the mean is greater than the mode. This situation is also called
positive skewness.

Skewness [Image 16] (Image courtesy: https://www.safaribooksonline.com/library/view/clojure-

for-data/9781784397180/ch01s13.html)

How to the skewness coefficient?

To calculate skewness coefficient of the sample, there are two

methods:

1] Pearson First Coefficient of Skewness (Mode skewness)

Image 17

2] Pearson Second Coefficient of Skewness (Median skewness)

Image 18

Interpretations

 The direction of skewness is given by the sign. A zero means no

skewness at all.
 A negative value means the distribution is negatively skewed. A
positive value means the distribution is positively skewed.
 The coefficient compares the sample distribution with a normal
distribution. The larger the value, the larger the distribution
differs from a normal distribution.

Sample problem: Use Pearson’s Coefficient #1 and #2 to find the

skewness for data with the following characteristics:

 Mean = 50.
 Median = 56.
 Mode = 60.
 Standard deviation = 8.5.

Pearson’s First Coefficient of Skewness: -1.17.

Pearson’s Second Coefficient of Skewness: -2.117.

Note: Pearson’s first coefficient of skewness uses the mode.

Therefore, if frequency of values is very low then it will not give a
stable measure of central tendency. For example, the mode in both
these sets of data is 9:
1, 2, 3, 4, 4, 5, 6, 7, 8, 9.

In the first set of data, the mode only appears twice. So it is not a
good idea to use Pearson’s First Coefficient of Skewness. But in
second set,

1, 2, 3, 4, 4, 4, 4, 4, 4, 4, 4, 5, 6, 7, 8, 9, 10, 12, 12, 13.

the mode 4 appears 8 times. Therefore, Pearson’s Second

Coefficient of Skewness will likely give you a reasonable result.

Kurtosis
The exact interpretation of the measure of Kurtosis used to be
disputed, but is now settled. Its about existence of outliers.
Kurtosis is a measure of whether the data are heavy-tailed
(profusion of outliers) or light-tailed (lack of outliers) relative to a
normal distribution.

Kurtosis [Image 19] (Image

courtesy: https://mvpprograms.com/help/mvpstats/distributions/SkewnessKurtosis)

There are three types of Kurtosis

Mesokurtic

Mesokurtic is the distribution which has similar kurtosis as

normal distribution kurtosis, which is zero.

Leptokurtic

Distribution is the distribution which has kurtosis greater than a

Mesokurtic distribution. Tails of such distributions are thick and
heavy. If the curve of a distribution is more peaked than
Mesokurtic curve, it is referred to as a Leptokurtic curve.

Platykurtic

Distribution is the distribution which has kurtosis lesser than a

Mesokurtic distribution. Tails of such distributions thinner. If a
curve of a distribution is less peaked than a Mesokurtic curve, it is
referred to as a Platykurtic curve.
The main difference between
skewness and kurtosis is that the skewness
refers to the degree of symmetry, whereas the
kurtosis refers to the degree of presence of
outliers in the distribution.

Correlation
Correlation is a statistical technique that can show whether and
how strongly pairs of variables are related.
Correlation [Image 20] (Image courtesy: http://www.statisticshowto.com/what-is-correlation/)

The main result of a correlation is called the correlation

coefficient (or “r”). It ranges from -1.0 to +1.0. The closer r is to
+1 or -1, the more closely the two variables are related.

If r is close to 0, it means there is no relationship between the

variables. If r is positive, it means that as one variable gets larger
the other gets larger. If r is negative it means that as one gets
larger, the other gets smaller (often called an “inverse”
correlation).