How to read a Boxplot?
In a bid to make our statistics more accessible, we thought of putting out information on some of the graphs/charts we use. Here’s the first in such series.
What is a Boxplot?
A Boxplot shows the distribution patterns of data. The strength of a Boxplot over other plots is, it shows the central tendency of the data, it helps to understand the shape of the distribution as well as it shows the whole range of observations.
The grouped Boxplot even helps compare between different groups in a dataset.
The above grouped Boxplot shows cases for different ITAT forums, disposed during financial years 2010 to 2017.
Also read: The Burden of Disputes!
Let’s take the example of ITAT Ahmedabad in the above box plot for a deeper understanding.
A. Of all cases decided for Ahmedabad, the blue box represents the time taken in years for the mid-50 percent of cases, here it spans from 2 to 4 years. The horizontal line inside the box indicates the median. Median is the 50th percentile. This means 50 percent of cases took more than 3 years for disposal in Ahmedabad location.
B. The horizontal line below the box represents Q1, first quartile. This is 25th percentile. This means about 25 percent cases took less than 2 years. This distribution is represented by D.
C. The horizontal line above the box represents Q3, third quartile. This is 75th percentile. This means about 75 percent cases took less than 4 years. This also means about 25 percent of cases took more than 4 years to resolve. This distribution is represented by E and F together.
F. The circles represent cases that are outliers, cases which do not follow the normal trend for that particular group.
(Written by Rutuja Udyawar, VP Analytics @ Riverus.)