Assignment

You are asked to carry out an analysis of a dataset and to present your findings in the form of a maximum of two (2) visualisations, (or 2 dashboards comprising a set of linked sub-      visualisations) along with an evaluation of your work .

You should find one or more freely available dataset(s) on any topic, (with a small number of restrictions, see below) from a reliable source.

You should carry out an analysis of this data to determine what the data tells you about its particular topic and should visualise this data in a way that allows a user to understand the data and what the data shows. You should create a maximum of two visualisations of this  data that efficiently and effectively convey the key message from your chosen data . It         should be clear from these visualisations what the message from your data is.

You can use any language or tool you like to carry out both the analysis and the                      visualisation, with a few conditions/restrictions, as detailed below. All code used must be   submitted as part of the coursework, and it must include enough instructions/information to be able to run the code and reproduce the analysis/visualisations .

You should create a very short (2 page, ~800 words) reflective evaluation of the success (or not!) of your completed visualisation(s) in terms of the visualisation principles discussed in   class. You should comment on their strengths and weaknesses, and any improvements you    would like to make to them in future . This evaluation should critically discuss the                      visualisations you have created. It must *not* just be a description of what the                        visualisations show and what they tell us about the data . It should cover *why* your            visualisations do a good job of communicating the information that you have found out        about your data.                                                                                                                                         

Tool usage

Although you are free to use any tool, language, library that you like, there are some exceptions/conditions to this for you to be aware of.

In order to mark this assessment, I need to be able to see it! You absolutely *must* submit absolutely everything that is needed to create and view your assessment. This should          include all code used to clean and filter data, any intermediate datasets generated, and so on.

You must include enough instructions/information to be able to run the code and reproduce the analysis/visualisations.

Tableau

If you use Tableau to create your visualisations, you *must* ensure you are either creating dashboards that combine multiple sub-visualisations together with some form of linked   functionality between the sub-visualisations on each dashboard, or an effective story         presentation. Simply creating individual non-linked visualisations will not suffice .

If you use Tableau to create your visualisations you must submit a packaged tableau workbook that includes all needed resources within the packaged .twbx file 1 .

PowerBI

If you use PowerBI to create your visualisation, please make sure it is possible for someone to view your submitted visualisation. I don’t know what you need to do with PowerBI to     enable this, but I do know that I have previously had PowerBI submissions that I have not   been able to view correctly. You must submit a .pbix file that I can open in PowerBI, and     preferably a link to the online visualisation.

Python/JavaScript/…

Please submit a list of all libraries required to run your code/visualisations. This might be a pipfile or a requirements.txt for Python, or a package.json for javascript, and so on.

Java

No. No Java. It’s the only programming language that’s banned. I just can’t deal with the classpath issues.

Dataset Selection

You are free to choose data on any topic you like, with the following exceptions. You cannot use data connected to the following topics:

1.   COVID- 19. I’ve seen too many dashboards of COVID-19 data that just replicate the   work of either John Hopkins or the FT, and I’m tired of seeing bar chart races of        COVID deaths, which are incredibly distasteful. Let’s not make entertainment out of a pandemic.

2.   World Happiness Index. Unless you are absolutely sure that you’ve found something REALLY INTERESTING that correlates with the world happiness index, I don’t want to see another scatterplot comparing GDP with happiness. It’s been done too many      times.

3.   Stock Market data. It’s too dull. Treemaps of the FTSE100/Nasdaq/whatever index you like are going to be generally next to useless, candle charts are only useful if   you’re a stock trader, and I don’t get a thrill from seeing the billions of dollars         hoarded by corporations.

4.   Anything NFT/Crypto related. It’s a garbage pyramid scheme that is destroying the  planet and will likely end up hurting a bunch of people who didn’t know any better.

Learning Outcomes Assessed

1.   Describe and discuss the theory behind visualisation design

2.   Critically analyse visualisations of data

3.   Examine and explore data to find the best way it can be visually represented

4.   Create static, animated and interactive visualisations of data

5.   Critically reflect upon and discuss the merits and shortcomings of their own visualisation work