Data Analysis Fundamentals

The world runs on data. We continually generate, collect, and discuss it. A truly enormous amount of information is constantly being produced and stored by countless tools, organisations and individuals. All of this is gathered in different ways and stored in different locations; frankly, it's a mess. The role of a data analyst is to work with these vast fathoms of data. To collate them and make them consistent, to identify patterns and relationships, to test the significance and validity of ideas and proposals.

In this module, you'll be introduced to some of the key techniques, tools, and practices of a data analyst. You'll work with -- and even improve -- large and awkward datasets. You'll identify stakeholder needs and meet them through data exploration. You'll seek and find the truth.

Investigation

The real world is inherently chaotic, stuffed with irreducible complexity and nuance. This means that analysis -- useful, meaningful analysis -- is not a straightforward thing. Effective data analysts are those who can take a general request, refine it down to a specific question, and ruthlessly interrogate the data to establish not just an answer, but an understanding of how meaningful/reliable/generalisable that answer is.

Throughout the coursework, you'll be presented with several tasks. Many, if not all, of these will have hidden depths, requiring you to dig a little deeper, to not be satisfied with the easy path but the correct one. You should be relentless in your pursuit of evidence and meaning.

Communication

There's no point in doing analysis if it never gets used. No matter how much time you spend exploring a dataset, if other people can't access or understand your findings, your work will never have an impact. Even if you do communicate it, poorly-expressed analysis can lead to the opposite effect than intended/required.

An indispensable part of analytics is communicating your insight in an effective way. During this module's challenges, you'll be tasked with converting your in-depth understanding and exploration into a format that can be shared with stakeholders. You'll need to summarise your ideas, make them approachable/relevant to the audience, and find effective ways to stress key points.

Recommendation

A key element of analysis is being able to move beyond just finding information and blandly presenting it towards finding the right information to inform action. There are an infinite number of details you could highlight in any analysis, but so what?

There are nine million bicycles in Beijing and that's vaguely interesting for a second after hearing it and then means nothing to anyone who wasn't planning to open a bike shop there. In London, you are never more than 100 feet from a rat, but that fact assertion causes terror rather than bringing any benefit. Why are those details interesting? Based on what you have found, what should we do? How do we benefit from the analysis?

Analysts must not just find information, but find the right information. They need to find the 'so what' in everything that they do, to tie data to meaning to action. They need to give people a reason to care about their results.

Structure

This section of the course consists of a single two-week case study. You'll take on the role of a data analyst, researching and communicating insights to business stakeholders.

Case Study 1: FAA Warehouse (Weeks 1 and 2)

A manufacturing company wants to attract new US-based aviation clients, and needs data to do it. You'll work with an early-stage data warehouse, managing and exploring complex datasets on air travel in the USA.

This will involve heavy use of SQL, data cleaning & processing with Python, and in-depth, open-ended analysis tasks.