The Different Categories of Data Analytics

Data Analytics comes in different flavors. All of them are, at the core, crunching numbers and statistics to form a coherent narrative. The difference lies in what the narratives are used for.

Today, we will be examining the different kinds of Data Analytics that are used in the industry. For the purposes of relevance, we will be using the COVID-19 pandemic for examples.

Descriptive

This is by far the most common type of Data Analytics, and there is good reason for that - it is basic and provides insights to a current situation. Figures are displayed in a way that shows how the situation currently is. It describes a situation, hence the word descriptive. The question most commonly answered in these scenarios is what, which and possibly where or when.

Examining the
current situation.

One example of descriptive Data Analytics would be a day-to-day tally of new COVID-19 infections, with perhaps a monthly average and comparisons with figures from preceding years.

Diagnostic

The next step after descriptive Data Analytics is diagnostic Data Analytics. This is an analysis of the data, with possible patterns and correlations identified within. It is meant mainly to answer the question of why - why the situation is as presented and the possible causes.

Possible causes such as age
and pre-existing conditions.

One example, building on the example given above, are further breakdowns of COVID-19 data into age groups, genders and pre-existing medical issues. From this data, analysts can then determine if the numbers of infections or deaths correspond to these conditions.

Predictive

Progressing naturally from diagnostic Data Analytics, which explores how the current data came to be, is predictive Data Analytics, which projects, from the current trends and conditions, what the data will probably look like from this point forward. This is the Data Analytics version of fortune-telling.

Telling the future.

This could take the form of dashboards where variables can be adjusted to see what effect this has on the visualization. For example, if the number of infected elderly people was brought down, what effect would it have, over time, on the total number of COVID-19 infections?

Prescriptive

The final piece of the Data Analytics puzzle is prescriptive Data Analytics. After predictive Data Analytics, this comes in the form of recommendations as to how to affect the projected trend. As one might imagine, this is probably the most valuable component. The descriptive component explores the what, the diagnostic component generally explores the why, and the predictive component explores the what-if. The prescriptive component, however, adds value in the form of analyzing the gathered data to determine what steps could be taken to achieve desired outcomes.

Recommendations and
action plans.

In the case of our example, the desired outcome is lower, controllable and (hopefully) zero COVID-19 cases. Sample advice might be, vaccination programmes for elderly and high-risk groups, limitations on group sizes in public, and so on.

In a nutshell

What I have outlined is the different kinds of Data Analytics and how they stack up. The examples given above only pertain to the COVID-19 situation. But imagine these analytics being applied to profits, research and the stock market!

Analyze this!
T___T