How the UX Design Thinking Process is Applied to Data Visualizations in Data Analytics

With empathy in mind (to begin with); the data visualization in the opening image is great, except for one crucial thing. Can you figure out what that is? It is the use of Christmas colours together, of which is not ideal for those who are colour blind, and for users (in UX Design) and stakeholders (in Data Analytics) with any form of disability impediment. According to Colour Blind Awareness, over 350 million people worldwide are actually colour blind. (On the other hand, the exception to this rule would apply to visualizations' on world maps, for example with global CO2 emissions.) For those of you lovely readers who write for Vocal Media; you will be pleased to know that the bar chart in our writers statistics page is a great example of a Data Visualization gone right. Why? The y-axis starts at zero, because the number of readers reading your work is unknown at any given moment in time; and (furthermore) the bars are easy to spot trends, and most of all, to read.

Moreover, green is associated as a positive colour (no different to when it resembles go at traffic lights), while red is associated as a negative colour. If such a visualization was on trade per country; the country with its trade numbers in red could assume a trade deficit, of which could be further from the truth. Therefore such data could come across as misleading.

Both tech specialities (that is, Data Analytics and UX Design) agree that empathy is needed for their target audience; users and stakeholders again, depending on the context. Both consider the needs, pain points, and emotions of the people they need to serve. Both specialities have a human element when it comes to computerized interactions, and working with technology.

Data Analytics goes one step further by asking some quality questions before designing their visualizations in tools such as Tableau, Power BI, and Looker Studio (to name); and even on spreadsheets:

* Do the colours and labels make sense in this visualization?

* Am I selecting the right statistical representation for the data to be shared to stakeholders?

* Are the nine principles of design (from balance to unity) utilised?

Both tech specialities also keep empathy in mind when undertaking some further research. Data Analysts usually only conduct surveys, where the sample size is pre-calculated on their end as follows:

S= Z 2 × P × (1−P)M2.

In other words, if the size of the relevant population (say, of a city) is 500, with a 95% confidence level, and thereby allowing for a margin of error rate of 5%; Data Analysts/Data Analytics Consultants would survey 218 people in that relevant city at random, in collecting the right data to analyse, and then to visualize.

The define phase in UX Design is where any research is synthesised in defining the users pain and gain points; in other words the problem that needs to be solved. This problem is further defined by working on a number of UX deliverables from concept mapping, to creating value proposition canvases and user personas to name. In Data Analytics, this process is embedded into the entire Data Analytics thinking process as follows:

Ask - Prepare - Process - Analyse - Share (the data visualization stage), and then to act on the right and proper use of this data.

When it comes to creating the data visualization; those working in Data Analytics figure out what their stakeholder/s need to know from the data that has been analysed (and even collected through surveys). For example, if a coffee roaster wants to know total sales by territory, and the commissions made by each member of their sales force; that data gets sorted and cleaned first. These stakeholders do not have time to go through multiple lines and rows in spread sheets (usually small data); and a few SQL queries would need to be run in programs such as BigQuery to make sense of large data bases. Let alone hunt for the right public or private dataset instead.

In the ideation stage, UX Designers regard this stage as iterations. In Data Analytics, this is where Data Analytics Consultants/Data Analysts come up with ideas, and what dataset/s to use to share this data with the relevant stakeholders, through this visualization stage. Additional features for an existing dashboard (whether live or otherwise) may prompt the following questions to be answered to name:

* What new data visualizations would help the relevant target audience (users and/or consumers)?

* Would bar charts and line charts work, in addition to the donut chart?

The next stage is the prototyping stage. For UX Design, this is usually the sketch (lo-fi prototype) of a new app (for example), or the prototype wireframes in programs like Figma (hi-fi, and pixel perfect). In Data Analytics, this is the stage where the actual data visualizations' themselves are created in programs such as Tableau, or on spreadsheets, with or without pivot tables. Examples of data visualizations' can be found by clicking here.

That leaves the final step of the design thinking process, which is the testing phase. In UX Design, this means usually recruiting a sample of the target users, and running user testing sessions with them, in exchange for an incentive to thank them for their time. If you have ever tested out a prototype (a new app or otherwise) in exchange for a gift voucher or similar; then chances are a UX Designer would have been facilitating that session, while such feedback is sent to the developers afterwards. In Data Analytics, the testing phase when it comes to data, and data visualizations' involves doing a test presentation of your visualization to your colleagues, before presenting such to stakeholders; and in turn gathering any feedback (constructive or otherwise on such). If any iterations are required either way; they shall happen. The ultimate test is when the relevant data insights are shared with the stakeholders who requested such insights in the first place. The litmus test is to ascertain the speed at which they absorb and understand the data, and the data visualization being presented to them. And on how many probing questions they might ask. And whether or not the relevant stakeholders will act (the last stage of the Data Analytics thinking process) on the data presented to them, in getting the right Return On Investment (ROI).

It is actually quite mint, where the logic of UX Design thinking can be applied to the Data Analytics profession in creating intuitive and relevant data designs and visualizations'. You're welcome fellow data nerd.

NB: Justine wrote this article at the time of being a student in Data Analytics.