What is Descriptive Analysis?

There are many different types of data science analysis methods. Let’s start off with descriptive analysis, which is just what it sounds like. The objective of descriptive analysis is to describe a set of data. This is usually the easiest form of analysis, and it will often be the first type of analysis you do. In the simplest terms, descriptive analysis is when you just summarize your data, and it is not really meant for coming to conclusions about the relationships between variables. Some types of simple descriptive analysis strategies are calculating mean, median, and mode. These statistical strategies are called measures of central tendency, because these numbers can be referred to as the most prominent numbers depending on how you calculate them. For example, mean is a prominent number because it tells you the average of the data set. On the other hand calculating range, standard deviation, or variance are also descriptive analysis strategies, but they are called measures of variability. This is because they are numbers that tell you how diverse the data set is. An important note to make is that descriptive analysis is just meant for summarizing a sample. It is not for generalizing the data or figuring out if there is a correlation between variables. In science class, they always tell us that the data table should be hard facts, and there shouldn’t be any calculations or opinion. The analysis section of a lab report is where you get to interpret the data and perform those calculations. Similarly, descriptive analysis is just the first step, and making interpretations is later on the staircase, because it requires more methods of analysis. Now you may ask, why would we ever perform descriptive analysis when we can analyze further? The answer is simple, not all data needs to analyzed further. Sometimes, you just need to present information that may not even have a connection. A good example of when descriptive analysis is actually used is censuses. Censuses are surveys the government gives out to the country’s citizens, and it is used to collect general information about the country’s population, such as the age, sex, ethnicity, etc. The goal of a census is to just describe the population, not to make predictions about the future or make a generalization. Therefore, descriptive analysis is perfect for the task.

References: https://www.coursera.org/learn/data-scientists-tools/lecture/7Jmzs/types-of-data-science-questions