Statistics and data science

Consider the following quotations highlighting the main features of Statistics and Data Science:

“Statistics is the science of learning from data.” – Davidian, M. and Louis, T. A. 2012. Why Statistics? Science. 336, p.12.

“Data science is the extraction of knowledge from data.” – Diggle, P. J. 2015. Statistics: a data science for the 21st century. Journal of the Royal Statistical Society. 178 (Series A). pp. 793–813.

Thus, as scientific disciplines, both Statistics and Data Science focus on studying data.

But how do they compare? For instance:

• Is there an overlap and what are the differences?
• Do statisticians need to be aware of Data Science?
• And do data scientists need to bother about Statistics?

Use the comments sections at the end of the page to add your thoughts. 

The quotations above suggest that Statistics and Data Science have a lot in common.

But there are also crucial differences stemming from the words ‘learning’ and ‘extraction’.

• Statistics aims to obtain inference about possible statistical models behind the data, which help understand the data and quantify the uncertainty therein. Models are based on probability and allow for a meaningful interpretation of the fitted models and their implications.
• Data science aims to extract hidden patterns in the data that reveal its structure and help predict future values. Methods involve statistics but mostly rely on computational algorithms using machine learning and, more recently, artificial neural networks aiming to replicate the learning processes in the human brain.

Thus, the two sciences are not the same but rather provide helpful companions to one another by enhancing their computational, inferential and predictive powers.

Next steps

So far, you have been introduced you to the main concepts you will study in this course. Next, you will consider four case studies and the types of research questions the data may help to answer.