Summarizing data and EDA
Summarizing data and performing EDA is the first task that you, as an analyst, always need to undertake. It will help you to understand the data you are dealing with and know which techniques are appropriate. But EDA is more than just doing some pivot tables. We need to start with a primer on descriptive statistics.
Primer on descriptive statistics
Data can be qualitative (also known as categorical) or quantitative.
Qualitative data can be nominal when, for example, you have two choices, such as “yes” and “no” or “male” and “female,” but you do not have an implicit hierarchy in it. It can also be ordinal when a hierarchy has been implied, such as “level of education.”
The most common way to describe this type of data is through the use of tabular methods such as frequency distributions or graphical methods with proportions.
When working with statistical models such as linear...
