Counting Differently Seminar: Will Stahl-Timmins on ‘Visualising Health and Environment Data”

By Farzana Dudhwala and Thomas Turnbull

Last week Will Stahl-Timmins, an associate research fellow at the European Centre for Environment and Human Health (ECEHH), based at the University of Exeter, presented his thoughts on the visual presentation of data and information.

The Counting Differently Working Group in action

The Counting Differently Working Group in action

From the very first slide we knew we were in for a treat. Despite a flurry of recent interest, Will pointed out that effective visual representation was nothing new.  Will showed us a picture of what statistician Edward Tufte has argued is “probably the best statistical graph ever shown”: Charles Joseph Minard’s (1869) depiction of the losses suffered by Napoleon’s army in 1812 . This ‘statistical graph’ is often hailed as a lesson in how visual communication should be done:

Charles Joseph Minard’s depiction of the losses suffered by Napolean’s army in 1812

Will then showed us another example of presenting data visually. Florence Nightingale’s “Diagram of the Causes of Mortality in the Army of the East” (1858) is often cited as an  example of objective data communication. However, as only a graphic design practitioner could, he pointed out how this seemingly ‘objective’ visualisation was in fact tweaked to turn it into a rhetorical device.

Florinighting

In Nightingale’s diagram, people often assume that the area of the wedge is proportional to the figure it represents, but this is not the case. In fact, the figures are represented by the radius of each piece of the circle – not nearly as intuitive (at least not for people in cultures where the area is the intuitive conclusion rather than the radius). The effect emphasises the extent of disease driven fatalities, at the expense of combat fatalities. Nightingale’s Rose, as the diagram is sometimes known, can certainly mislead; though of course in her case this was done with the best of intentions.

After an informed introduction to the world of visualisation, Will put on his other hat to speak about the role of data visualisation in academia. Often, he explained, his work for the European Centre for Environment and Human Health involves dealing with what he calls ‘excited scientist syndrome'(!). This is the entirely understandable enthusiasm of the expert scientist when they feel they have come across an exciting data-driven insight. Will’s job is to visual render the enthusiasm of the expert in such a way that the non-expert can become equally excited.

So, when Will received a visit from a colleague suffering from excited scientist syndrome, he was able to translate their enthusiasm about the – as yet undocumented – relationship between climate change and pharmaceutical use into an effective and informationally rich indicator (the paper that came out of this can be found here, and the resulting graphic can be seen below):

CCandPHARMA_v35

Will also spoke about the ‘Pharma Town‘ project, an attempt to map the flows of pharmaceuticals through the environment. A visualisation he created in the middle of a study acted as a focus point, around which his fellow researchers could direct their expertise, and point out where probable diffusions and flows might, or were known to, occur. They used the visualisation as a tool to think about new ideas and ways of sharing knowledge. In other words, as Will told us, the visualisation revealed hidden connections in the work and allowed them to be brought to the fore. This is very much in line with Michael Lynch’s argument (see below for a link to Lynch’s paper) about how pictures can “transform  previously hidden phenomena into visual displays for consensual ‘seeing’ and ‘knowing’” (1988:203).

pharmatown

In another example, Will showed how reams of spreadsheet data can be transformed into icon driven and aesthetically pleasing images, which can convey some of the excitement that the scientist feels in their own work. The images also work as a useful visual heuristic for the scientists themselves. So, for example, a large quantity of clinical trial data can be summarised visually, digested rapidly, and, in some cases, remembered more accurately. Will is looking to automate the visualisation process, which would no doubt be of benefit to busy health professionals.

The flip-side of the visualisation process is assuring that the exciting this rendering still accurately portrays the science which it communicates. So Will’s relationship with the science he translates is based on trust, and his own understanding of the research he conveys.

Further academic robustness comes from the fact that Will and his colleagues are work to empirically test the effectiveness of information visualisation. Using relatively large subject samples, they have empirically tested how effective various forms of visualisation are at appealing to, communicating with, and sticking facts into human brains. Their initial findings seem to point out that for certain demographics visualised data is more appealing, more quickly digested, and better remembered.

Will’s talk, as hoped and expected, sparked a number of interesting questions and discussions. Here is a selection of those that were particularly thought-provoking:

  • Does the stage at which a visual representation is produced affect the way the researchers themselves relate to their own data? Does it change the way the report or journal is written or how the conclusions are drawn?
  • Are there cultural differences between the way the designs of the visual representations are received by their audiences?
  • To what extent is it really possible to represent risk and uncertainty?
  • Is a picture more misleading than textual prose because a packaged picture seems to convey some sort of higher truth (much like arguments that have been made in STS about numbers)?
  • Do people that make infographics or visualisations have to provide a methodology? Does the creator of the visualisation have to provide a record or justification for what gets counted or what doesn’t get counted in the infographic?
  • What would peer review look like for visual science?

Will also kindly gave us a resource list to draw upon if we wanted to find out more about how to create data visualisations ourselves, or if we wanted to see some examples of other people’s visual work.

The two programming languages he mentioned are:

Some useful websites are:

And  some conferences too:

And lastly, In case you missed it, we had suggested two readings for the group which are particularly relevant to Will’s talk:

The group will meet again at 12.30pm on November 29th at the Said Business School in Seminar Room 13, where we will welcome Dr Anthony Garnault from the Department of Contemporary China Studies at the University of Oxford. He’ll be talking about social surveys in the twentieth century, and will be questioning what the ‘typical’ represents. We’ll be sending out a couple of suggested readings shortly, so watch this space! (A full list of seminars and activities can be found here).

The (eighth) STS Talk-Walk: Visualising – what is it to visualise?

by Andreas Birkbak

The eighth installation of the STS Talk-Walks revolved around the theme of visualisations and visualising. What practical work goes into creating visualizations and making them travel? Who, which or what is actually being visualized? What are the risks involved in visualizing? How come that ‘visualization’ has gained such currency? Has it, actually?

Whether it was due to the all-too-familiar visualisation of ‘heavy rain showers’ on the BBC Weather homepage or not (the grey icon almost scared this writer off), the turnout was not overwhelming this time. However, the happy three that ventured out in the rain were rewarded with great conversational depth and a very tolerable amount of rain. Here follows a necessarily incomplete and biased story of how it went, but really this image says it all (or does it?):

A talk-walk visualization.

Walking downstream on the Thames Path from Said Business School, we started by noting how the term ‘visualisation’ is claimed by a confusing collective of actors. It is taken to mean anything from diagrams that sum up texts or ideas to digital stepping stones that have become integrated into how scientists proceed in laboratories. It was suggested that while this diversity makes it hard to run smooth conferences under the visualisation banner, it could also be cast as a powerful unifying concept. Further to this, one participant pointed out how a visualisation in itself can work as a focal point around which people with different interests are able to gather. Interestingly, our shared everyday language also seems to contain a throng of hints to the visual, as is for example ‘illustrated’ when politicians talk of the need for ‘transparent’ institutions and ‘clear’ policy making.

Later, a halfway pint at the cosy Isis Farm House became instrumental in discussing the concept of affordances and the ontological status of those environments that present themselves visually to us. None of the participants were entirely comfortable employing the notion of affordances (which has also been used by a very diverse set of writers), but it was useful for trying to get at what it is that might make visualisations appear so powerful in comparison to for example texts. Perhaps there is something more immediate about the way that we relate to the visual? This argument was only hesitantly accepted, though, with some unexpecting readers suggesting that they reacted in a highly emotional and instinctive way to the final passages of the novel ‘The Road’ by Cormac McCarthy. In this case the text seemed to have powers equal to the most arresting (in the Barthean sense) pieces of visual art.

During the walking and talking, we several times found ourselves reflecting on the recent InSIS conference Visualisation in the age of computerisation. For example, the keynote by Peter Galison was useful for thinking about how different styles of attaching truthfulness to scientific visualising have emerged over the last three centuries. The idea came up that if digital visualisations are gaining currency in our times, it might have to do with the classic statistical logic of attaching value to representations that sum up large amounts of data.

It turned out that more than one of the talk-walkers were doing current work in which they encountered the practicalities of visualizing, so this occasion to discuss was warmly welcomed. Whether some of the thoughts that were generated at the talk-walk can be transferred from the field of the Thames Path to the field in which visualising work is carried out remains to be seen.

Next STS Talk-Walk: Friday, 22 July 2011.

Thank You: Visualisation in the Age of Computerisation

by Timothy Webmoor

“This conference should have been three days.” Comment overhead (by Anne Beaulieu) at a lunch.

Thanks to everyone who participated in the recent Visualisation in the Age of Computerisation conference over the weekend at the Saïd Business School.  Numbers are by no means the only measure of success. As a run down of the event, however, there were:

  • over 130 participants from North America, Australia, Europe and the UK
  • 50 paper presenters
  • 10 artists/scholars with installations and posters
  • 15 session chairs
  • 3 keynoters
  • 2 summarising disscussants

We also had a lively presence in ‘Twitterscape’ at #oxvisual. Some tweets we received:

Early photographer Stieglitz: ”In one’s way of seeing lies one’s way of action” – art challenging scientific ’objectivity’ at #oxvisual

Been looking forward to Mike Lynch’s talk all week – the geek in me is already satisfied at the pic on the title page of his ppt. #oxvisual

We will shortly be updating the conference website with links to the video of the keynotes. We welcome comments at this blog, tweeted at #oxvisual or sent to the organisers at oxfordvisualisation@sbs.ox.ac.uk. And please do send photos or links to photos.

Finally, thanks again to the organising team, our sponsors and to the Institute for Science, Innovation and Society for hosting the event.

Visualisation in the Age of Computerisation conference: Programme and registration

by Timothy Webmoor

The Institute for Science, Innovation and Society (InSIS) is organising a two-day conference on 25-26 March 2011 at Saïd Business School, University of Oxford, with support from the Norwegian University of Science and Technology (NTNU), the Oxford e-Social Science project, Digital Social Reserach, eResearch South and CD4.

Date: 25-26 March 2011
Location: Saïd Business School, University of Oxford (map)

About the conference:

The theme of the conference is the permeation of science and research with computational seeing. How does computer mediated vision as a mode of engagement with information as well as with one another affect what we see (or think we see), and what we take ourselves to know?

Background and themes
Conference programme
Keynote speakers
Travel arrangements

Register online

Speakers include:

Peter Galison, Department of the History of Science, Harvard University
Michael Lynch, Department of Science and Technology Studies, Cornell University
Steve Woolgar, InSIS, Saïd Business School, University of Oxford

Summarising discussants:
Anne Beaulieu, Virtual Knowledge Studio
Paolo Quattrone, IE Business School and Fulbright New Century Scholar

For more information:

www.sbs.oxford.edu/visualisaton
or
Contact visualisation@sbs.ox.ac.uk