This blog records and reflects upon the challenges, successes, and failures of creating data visualisations using Tableau Public, Data Hero, and The World Bank data visualisation tool. 

These tools were used to create data visualisations to illuminate a narrative exploring the parallel degradations of human health and the natural environment. The finished article can be viewed here.

All datasets and databases are fully referenced in the bibliography.

Comparisons Continue

The same issues arise with most charts now. If the Tableau charts take up too much space, and their higher functionality is irrelevant to their message, then what is the point? A reader shouldn’t be expected to have to click on a data visualisation to see it properly, that defeats the purpose, and in our ‘read-fast-move-on’ digital culture, not having the data on instant display is criminal.

So then it appears I have answered my own quandary; visual clarity over aesthetics and techy interactivity every time. It’s a hard-learned lesson for me, the point of data visualisation is exactly as it is named, to visually represent data, if it doesn’t do that effectively then it’s just not doing its job.

This visualisation is clear, obvious, and functional. It does exactly what it is supposed to do. It isn’t interactive, but it really doesn’t need to be.

Data Visualisation Comparisons

After creating some visually compelling interactive data visualisations with Tableau Public, they are embedding into my article, but the width of the area is less than needed for many of them to display correctly without clicking to view full screen. While not a huge issue, this certainly isn’t ideal, and now I’m left questioning the need for interactive data visualisations at all. The visualisation should be clear enough in its static form to show the data and its patterns, so after returning to Datahero, I am now left with a choice between the fully embeddable and arguably more visually attractive chart, and the clearer, yet static chart. There is the option of screenshotting a full-screen Tableau chart, but that doesn’t have the clarity of the Datahero one below. Decisions decisions….


I woke up with a breakthrough!

After slogging and struggling with trying to see correlations in my data, I realised I have been going about this all wrong. The whole idea behind this project is interconnection, so allowing the divisions of nations to cause problems with displaying data is part of the problem. It makes considerably more sense to look at worldwide totals over a period of time to see if there are parallel degradations in human health and the natural environment. I was trying to see correlations by looking at just the current situation, but that was never going to show any decline, just what the most recent data shows; a static representation.

Of course, this would happen after I had submitted the thesis, but there is always time to improve on a digital artefact.

My latest data viz, before this breakthrough. Still not good enough.

Yet More Tableau Data Viz

More Data Viz

Tableau Data Visualisations


The issues with embedding from Tableau are annoying, and have been irritating me. I know how much time these things can suck out of the limited time-receptacle that contains thesis production, so I really want to fix this asap.

My feeling was that this is an issue of vs There is very little option for customisation of beyond the themes and options provided, and there is no facility to adjust CSS, without paying a fee, and frankly, if I’m going to pay for the privilege of tweaking the CSS I’d much rather pay for full hosting instead of just an add-on.

So, I decided to check it out by posting the embed code to my academic website, also WordPress, but hosted by Reclaim Hosting. It worked perfectly, fully embedded and interactive as it should be. Great. Ish.

Screenshot (247)

No silly arrow, no display issues, link here:

The question now is what to do? Do I just pay for hosting, or persevere with a less-than perfect product which fails to do one of the most basic things a website should do?