i strongly agree. i am facing a science exchange next week where i meet with biologists and psychologists. and everyone has to present his or her topic in 4min. should get interesting^^.

You could get some inspiration at FameLab how to present the research at a basic level but still in a comprehensive and valuable way.

In computer sciences it is even more dramatic I would say. I have witnessed presentations with tons of formulae on the slides and everyone in the auditorium buzzed away after 10min (the talk lasted 45min....). This is not was it should be. Of course, there is some complexity but to transfer the overall idea, new and inovative ways should be use. maybe something like PechaKucha (http://www.pecha-kucha.org/). I attended some presentations and I was really diverting and interesting.

http://www.famelab.org

Our organization is delving deeply into this question of storytelling and science, and specifically how to get people from all levels of education and society engaged in meaningful dialogue about important environmental issues. We have developed a website for free access to spatial data where people can view, upload, download, query and analyze environmental data. Our hope is that people will begin to ask their own questions, form their own opinions and influence environmental policy in their own regions. This has been an especially effective way to communicate climate data, which is so relevant in this age, but so enigmatic as well. It helps for people to be able to view maps, pictures and articles about past climate and future projections. Take a look at databasin.org.

Very interesting applications from datavasin.org. I've just replay to my environmental colleages in some GIS Labs at Universidade de São Paulo (Sao Paulo University, Brazil). Analytical an Multimedia GeoVisualization is a very important issue nowadays !

Digital designing of visual languages that are issue (discipline) specific may be an important pattern for future communications. If we are educating people for 'scientific' visualization (or any kind of authentic inquiry) then, analysis and recognition of qualities of visual languages becomes a key issue of concern.

Great, we are going into several directions here. Let me try to summarize.

One dimension is the overall purpose, i.e. communication with other scientists but from a different domain. Thomas provided some good pointers here, although I have the impression that these will mainly attract young researchers. Wendy then gave this another twist, by including the communication with layman with even more diverse backgrounds. The presented web site looks interesting, yet I am wondering if you also actively support user engagement.

@Wendy: do you also facilitate your target groups in their participation, i.e. do you have awareness raining campaigns or is there any form of direct contact and follow up with individual contributors?

Pretty close to the initial question, the above considers the communication of science to others, whereas Xavier added the diverse relation, i.e. following the notion of crowd sourcing and also accounting for scientific work by citizen/layman. I did not think about this initially, but agree that this adds an interesting novel perspective. Whichever representation you choose, it should embrace an option for providing feedback and to include external findings – maybe in kind of mash-ups?

Mentioning multimedia is also a good point (from Carlos this time). Especially when dealing with multifaceted data. If we use 3D, motion, sound, and maybe more, it will become easier to communicate the required complexity.

Lev’s comment is indeed optimistic. I could buy into the statement that, if we have a universal representation (may I say conceptual model and encoding?), then visualizations should follow on their own. However, after spending the last decade on conceptual schema languages, semantic interoperability and data translation, I have my doubts about universality. There have been great works, for example on image schemas, but all the diverse purposes of data usage will ultimately require specialized forms of representation. There might very well be some underlying common grounds, which might lead to powerful new forms of visualization, but I am sure that that won’t be sufficient for each and every case. In this sense, I fully agree to Dean’s comment on discipline specific visual languages – maybe it is even discipline and purpose/audience specific. I would be more than happy if Lev proofs me wrong on this one!

In addition to the above, any feelings about current ways of handling this topic in your domain? In the geospatial sciences for example, many researchers are used to communicate with rather complex maps and claim to communicate well with their intended audience and even with citizen. Asking these people, it turns out that most presentations are still far too confusing.

Science is like a movie, without good presentation and visualization it become boaring. It very important in science to present in such a way that readers enjoy reading your discovery. Poor presentation is responsible for making our education and science boring.

Ultimate aim of science is not its mumbo-jumbo for hi-fi, but its simplicity,attractiveness and usefulness to the end-user. Teaching, training and extension technology must have both art and science components so that even most ordinary people can take maximum advantage of research achievements. I agree on having greater focus on this part of research.

In this competitive world, you have to sell your product. Scientist's product is science. To survive and get the grants funded in these crunch days you and your product must have a good visibility. Good visibility can only come with good advertisement and publicity and there by the impact increases i.e. end users appreciate the product. In olden days the the students themselves are the brand representatives of their mentors. Now mentors or the scientists themselves have to make their work more attractive to the end users. so one has adopt all techniques of selling. That is why the saying " Not only know how but also show how" is the right choice when so many distractions are available to the end users.

The Institute for Information Systems at the University of Applied Sciences Northwestern Switzerland is since some months strongly going into the direction of visualizing science by big pictures. Because I have a very long background in that direction (master in economy and master in graphic design) I was hired to develop these approaches on how to break complexity by telling visual stories. At the moment we are offering consulting in that direction, called iconomics (icons in economy), which was a invention and research topic of my Master in Design. There is also a new founded company in Basel, which is going very strongly into the direction of depicting complexity by reflective and iterative content generation. For more information, don't hesitate to contact me.

Here are some links: http://tinyurl.com/cquaqpk and http://blog.vizbiz.biz (sorry they are in german)

P.S: In Switzerland you have even a "anti powerpoint party". :-)

As scientists continue to work with increasingly large data sets, data visualization has become an incredibly important asset in their tool belts. This is especially vital when they must communicate across interdisciplinary lines with other scientists or when promoting their science to the public.

Being entertaining (as Gajendra puts it) is one dimension - also to attract the next generation - but isn’t it more critical to tell the right message to the right people? Scientists should in general take responsibility to reveal recent and upcoming challenges, propose/recommend solutions, and advise decision makers. I am pretty much with Rakesh on this one: establishing an attractiveness and usefulness are key, and both can be achieved if we find simple ways of communication. But at the same time, we should avoid oversimplification, for example every scientist is responsible to also indicate uncertainties in the messages her/she gives. Seeing all of this as marketing productions might be dangerous, especially because marketing people do not like to talk about uncertainties or errors (at least not those I know). Maybe a related question could be: (how) should we market science products?

Good to hear the news from Darjan, are you also applying Deep Zoom (one of the approaches I mentioned in the introduction)? I find this pretty powerful and would like to get your opinion on that. Could you maybe also say a few words on which kind of people your consult (or intent to consult in near future)? I would appreciate some statement on the concrete science domains. (This goes not only to Dajan, but to everybody, of course ;-) By the way, thanks also for the links!

Linda, it is great hearing about your activities. Do you also teach visualization techniques for reports/briefings (or know people who do)? I am particularly asking because of the large difference of preparing a graphic for peers (other scientists), for a superior manager, a politician or a newspaper. I strongly believe all are important, but how much of this is important to know for which type (?) of scientist? Visiting Muhammad’s comment, I would then consider the use of visualization techniques for the purpose of (data) analyses as yet another story, excellent.

Big thanks for all the answers so far. This is great stuff and going into many more directions than I expected… I hope that this continues for a bit ;-)

Wow, This is so far away from the original question you addressed to me, but we have all levels of engagement from discussion groups to working groups to user-contributed stories about how they are using the data and the site for their research, education, decision-making, storytelling etc. We have training videos, and people who give live help and workshops.

Is it far away? Initially - at least how I intended the question - this was addressed to scientists from any domain, in order to find out if visualization and storytelling is considered important in 'their' daily work. Although it is obvious, I did not think of those people who are actually working primarily on teaching those capabilities to others. Accordingly, I was very happy when seeing your initial answer and now I am interested to get to know more ;-)

Thanks a lot for pointing into this direction! We (in the geospatial informatics) need more of your kind, too many people think ONLY in classical maps to be delivered via the Web!

Hi Linda, thanks, also for the link! It would be interesting to hear more about your visualization strategy. Do you have a pointer for this one, too? It seems quite disconnected indeed, but who knows... maybe it raises some new and helpful thoughts.

We've just published the results of a survey we did of users of data visualisation applications (as part of the research for our recently launched immersive 3D visual analytics application) and that specifically asked - amongst other questions - how much people were using data visualisation for "story telling" as against more "quantitative" uses. The results for that section in summary were:

88% always or often looking for trends

63% always or often looking for clusters

78% always or often looking for anomalies

72% always or often looking for cause & effect

55% always or often looking for stories

Respondents came from a mix of business and scientific users.

You can download the whole survey report at http://www.daden.co.uk/media/white-papers/download/.

David

The reality of all sciences, and pretty much every other domain that requires communication, is that /everybody/ uses visualization and storytelling. Unfortunately, many of them don't realize this, and many of them take an actively hostile attitude towards the suggestion that they are applying these techniques.

Nonetheless, nobody looks at raw data. Plotting a graph from a table is visualization. Choosing a transform (log/log, etc) of the axes is storytelling. More effective storytelling and visualization communicates the data better.

Eventually, all sciences are going to have to come to grips that data does not communicate itself, and that scientists necessarily are involved in producing the communication, and in shaping the knowledge that is transferred by that communication. Pretending this doesn't and shouldn't happen is a dead end road.

Your fields are, while undoubtedly not doing everything that they can to communicate effectively, but by virtue of the fact that you already recognize that your data is spatial, and has a natural visual representation, you're already doing far better than many other domains. Try telling an old-school microbiologist that people will understand his data better if he does a better job of visualizing it and helping it tell the story he's trying to convey, and he'll tell you that his data speak for themselves, and that if people want to understand, they'll look at the data on his poster and understand it themselves. The fact that you're even thinking about the issue puts you worlds ahead in the game.

@David: Thanks for the pointer. I will certainly have a closer look, also to understand which kind of people participated in your survey.

@Linda: Great, please let me know when it is ready ;-)

@William: Good point, indeed most of us constantly apply visualizations for multiple purposes. Unfortunately, many of us with rather little understanding of the appropriate representation forms - depending on audience. Even in the geographic information sciences, where courses on geospatial statistics, data quality and digital cartography are rather common, I still see quite little material about story telling, using the right level of abstraction, and preparing the appropriate documentation for decision makers. Partially the latter comes with work experience, but it would not harm to incorporate some of this into common university courses...

The art of Painting an image using a verbal description is one of the most difficult

thinks I have ever attempted.

Frequently the results of your best efforts, using very detailed and long winded descriptions of processes produce obviously undesired results from the recipients.

If you think about what you are doing. It is a language translation from a visualization

in your mind to a written language. The receiver than takes that written language translation, and re-translates it into a personal visualization that is influenced by

his or her prior personal experiences.

The initial image in your mind, and the final image in their mind are very different.

If instead, you can add pictures to your presentation, then the initial image

that you project in your mind will be closer to the final image in the recipients mind, but it will still not be identical, as both make additional interpretations

based on prior experiences.

This might be a too simplistic view. Certainly, a picture tells more than a thousand words, but this is largely due to the complexity that you can put into it with graphical elements, text, color etc., and the way in which they offer people to explore information, e.g. be zooming and rotating. Taking more advanced forms, such as 3D graphics, motion pictures and the like, this gets even boosted - as long as it is done well. Still, some visualizations require additional explanation.

I strongly doubt that all humans think in pictures only. For example, many persons which think of numbers as colors, but you cannot generalize over it. In consequence, no matter if you provide information in written form or in a visually more advanced matter, you do not have any guarantee that the the intended interpretation is communicated unambiguously.

Just discovered this page and TED talk about the beauty of data visualization:

http://www.informationisbeautiful.net/2010/the-beauty-of-data-visualization/.

Very relevant!

Data visualization is incredibly important as a tool of communication - whether for hard or softer data. We used our clinical data on biobanking to link data on cancer with behavioral data, geographic data and environmental risks data to create a software prototype "User Gateway" so users (scientists and policymakers) can generate hypothesis on what might be linked to what and to allow users to ask questions about combinations of risks to develop scientific proposals.

Commercially JMP, the stats company, has been developing beautiful visuals to show quantitative data analyses.

The Polis Center at IUPUI developed and hosts SAVI http://www.savi.org/savi/ where they took data on urban living, environment etc and set it up so non-scientists, policymakers, new arrivals to Indianapolis, could use. They add health data as it becomes available. It is so nice to use that users dont realize they are applying science!! The other side to story telling! Because of the very severe cuts in funding and criticisms of universities and research progress, I think it is very important for us to communicate our science to non-scientists. Interesting and effective story telling and analogies are the easiest way to do this.

Nice topic. Story telling is an art and free form mostly used as entertainment in the arts. The framework of story telling can be applied very effectively to scientific data to make it intelligible to those learning and those who are not scientists.

Thanks a lot Janet, this is pretty much what I was hoping for! The examples you are providing sound great, let me have a closer look and come back to you in a few days!

Telling a story well requires art and creativity, but storytelling itself is actually deeply intertwined with how we perceive and remember things. As reluctant as some are to accept this, at a perceptual level, it's "what we all do", to convert visual information landing on our retinas, into memory and understanding.

Of course, as Sven alludes, there are always neuro-atypical exceptions, but for the significant majority of people, what is remembered out of what is seen, is directly connected to a mental "story" that we tell ourselves about the image. There's quite a lot of good research on this subject - reference "Change Blindness" or "Inattentional Blindness", particularly work by Daniel Simons (The Invisible Gorilla), for good information. For a simple personal example, look out your window for a few seconds, close your eyes, and try to remember a detail that you know must be there - for example, how many cars are parked along the street, or the number of house-plants on your neighbor's porch - but that you didn't consciously focus your attention on, and "tell yourself" to remember. You'll find that the information simply isn't in your head anymore.

Effective data visualization uses storytelling tools to help the viewer construct their "internal memory story", and to lead them to focus on the important features, so that what they see, understand, and remember, are the important facts that the visualization is trying to convey.

Ah, Werner, I beg to disagree...

Data does not communicate itself (or, at least, rarely does so effectively). While it's a noble and lofty goal to say that scientific visualization ought to communicate nothing but the data, and that it ought to be artifact free, it's an unfortunate fact of reality that the _consumers_ of scientific visualization are human beings. You simply cannot decouple scientific visualization from the human viewing it, and that means that it's going to be interpreted through the lens of a huge collection of personal, cultural and situational perceptual filters.

As such, if the goal is to _communicate_ something to the consumer of the visualization, it's an almost necessity to engage in "art", and in "non scientific" story telling - effectively producing a reverse-transform for the human perceptual lens - if you want to communicate most effectively. It would be a wonderful world if we could simply be purists and eschew responsibility for accommodating the viewer, but in reality this tact reduces the value of our science, in the name of protecting it. Any attempt at visualization that ignores these factors is going to do a less effective job at communicating the _science_ that it was intended to communicate.

Of course, I'm not advocating that we ought to use art and storytelling in scientific visualization to lead the viewer to believe untruths, but that it's our role to help them recognize and understand the truths - an even thinner tight-rope, if you will.

As an almost trivial example of the occasional necessity to "corrupt" the science if you want it to be understood effectively, consider image sharpening. Image sharpening adds _false_ information to an image, because adding that false information, makes the _real_ information easier for the human visual system to find and interpret. Do viewers acquire some false information from a sharpened image? Almost undoubtedly. However, on the balance, helping them recognize the true information by this corruption, is almost always a net benefit, compared to having them comprehend nothing, when the image has not been modified. Simply providing the original image and placing the responsibility for finding the truth on the viewer, is most unsatisfactory (though of course, the original should be made available so that the viewer can verify what they've perceived).

This, and numerous other filters in the human perceptual system act on, and influence a viewer's ability to understand a visualization even before the information starts to be processed consciously. After it's in the conscious domain, there are a host of additional learned personal, cultural and situational filters that additionally modulate understanding, and they need to be considered too, to maximize information transfer. Artists have been learning their way around this territory for millennia - we could do much worse than to learn from their examples.

I also think that data visualisation can be, some time, a trap for inexperienced people. It is a knowing fact that TV channels (imply visualisation news) can manipulate real stories just choosing some images from whole action.

Evan in science, the Storytelling and the Visualisation can be manipulative. Sometime, on conferences, a visual result is put under questions and discussed. But profanes can take the results as good. The images remain in their mind. Nonlinear equations, for example, conduct to very different results for very small changes of the parameter’s values. The visualisation results can induce manipulative results for the inexperienced audience.

You know: take care of what you see because not always is real. In science especially!

Interesting Werner -- it sounds like, in Astrophysics, visualization must have a role as a primary analysis tool?

I live somewhere between basic visualization theory, and the bio/life-sciences, and in the bio-world, visualization is primarily used for hypothesis generation. In that role things like sharpening and interpolation aren't considered problematic - so long as they're documented - because false positives are much less damaging than false negatives. A few type-I errors in interpretation can be tolerated, because they just generate illegitimate hypotheses that can be later rejected through more rigorous analyses, while any type-II errors in interpretation result in completely losing the most valid hypotheses about the phenomenon that's being visualized.

It's always interesting to see how different communities deal with these things. Most, I think, do it relatively poorly (biology most definitely included), largely as a result of a lack of comprehension of the human-viewer (and even some technology-based) components of a visualization. There's an overwhelming desire to "see the data as it is", but quite little grasp of what that actually means. Frequently, it seems that the best we can do is help them see what they need to see, regardless of what they say. The number of researchers I know who get uptight about contrast enhancement, but who don't know, or care, what the gamma correction for their monitor is, or whether it's calibrated, is depressing.

I could not agree more. Visualization is a part, which recently gains attention - at least in parts of my field (Digital Earth and Citizen Science) - but there are more mechanisms which help telling a good story in the end. So, who has some cases to share with us, e.g. on telling scientific insides to experts of another domain, politicians, decision makers, the public, family members, ...?

Science communication is very important and a largely ignored area.

Scientists tend not to be able to communicate there research to the general public or even governments and this is worrying for many reasons.

Firstly in the area of environmental science researchers are unable to use there finding on an national or international scale without the input of governments and if you cant commincate your reserch you have wasted that reserch.

Before you worry about storytelling and visualization you have to know the basics

Don't talk down to your audience

Drop the jargon

Don't force an issue

I believe this should be integrated into all PhD programs in an attempt to create a new generation of scientist that can communicate there research.

Indeed, the last two comments hit the underlying issue very well. Scientists (of all domains) should be very aware of the underlying principles and learn at least some basics of communication. Ultimately, the provision of scientific results (evidence) is our main business.

How this then translates into organizational structures is another story. Ideally, at least as far as I am concerned, a kind of communication department could provide guidance and support, but (i) such services are relatively rare, at least at the science level; and (ii) scientists might be unaware of such services or not pay enough attention to related activities.

I'm tremendously pleased to see the discussion come around to the issues of science and effective communication. Far too few scientists are willing to take responsibility for communicating their results, and instead try to take the position that their job is to find truth, or at least knowledge, and that it's someone else's job (perhaps the reader's) to extract understanding. I'll bet you I get another negative vote for writing this, but there's a pervasive attitude in the sciences that the scientist ought to be "above" interpreting the data, and that if the reader doesn't "get it", then it's the reader's/viewer's fault.

This is a tremendously appealing place to live - its just this side of impossible to be both the best communicator you can be, and the best scientist you can be, making it quite appealing to be the best at one thing, and hope that someone else does the other job for you - but as Sven and Jennifer mention, it's quite uncommon for that "someone else" to be readily available to and well-used by the scientists. As a result, we get poorly communicated science, and an impression that science is "talking down" to the lay reader (which is rarely actually the case - we just do a horrible job of communicating _that_ too).

Peculiarly, even in the visualization sciences, there is a reluctance to engage in "storytelling" with the visualization, to enable the visualization to communicate more effectively. To an extent, this is because the visualization sciences are still relatively nascent and are fighting for recognition amongst the "big boy" more established and traditional sciences. As such they are, I believe, beating the "We're Rigorous!" gong a bit too hard, and failing to take advantage of the opportunities they have to communicate better, in an effort to establish crisp/precise/rigorous/data/truth parity with the traditional sciences, amongst which they're still considered the weird new kid on the block.

I absolutely love seeing someone else saying that science students ought to be being taught visual communication tools - I'm going to quote you to my dept. chair - I think she's getting tired of hearing _me_ say that :-)

Thanks so much for all these replies. We went to a very interesting and important direction. Ultimately, being a brilliant mind and able to understand the world does not buy you (and others) anything, unless you can make people understand. There still is some tendency of packaging scientific results into cryptic sentences, and a certain fear of irrelevance in simplicity and clarity. However, also with the advent of citizen science, multi-disciplinary research and an increased demand for innovation, the scenery is slowly changing - for good, I believe.

Just an anecdote, We have been working for 3 years on software that maps clinical data on cancer patients and their lifestyle exposures with environmental exposures and geographic features in space and time. We wanted to bring new ways of looking at problems to epidemiology and public health scientists and policy makers. We have presented to many different kinds of audiences showing screen shots, diagrams, cartoons, text synopsis, etc, while we worked on the prototype. We are really interested in having a variety of users pick this up, but no-one shared our enthusiasm until recently. My colleague created a stunning animated visualization of the concept at our last presentation. There was an audible hiss of breath from our audience, whose response was "You've just solved two problems for us" as they launched a lively discussion of how they could enhance the software with their expertise to fit a new application. All it took was the one graphic slide for our audience to "get it", and for our communication to work. As a dedicated scientist and teacher, I think we have to persevere in trying to find the most effective communications, knowing that many humans, scientists and non-scientists, often find visualizations and stories the easiest way to understand even the most complex hypotheses and datasets.

Thanks Janet, this is an excellent example. I was just watching a two year old TED talk, which even takes one or several steps further. This was 'John Bohannon: Dance vs. powerpoint, a modest proposal' www.ted.com/talks/john_bohannon_dance_vs_powerpoint_a_modest_proposal.html

Absolutely recommended!

Sven: what a great video - thanks for posting.

How important? Very important. Teaching ethics in science requires a lot of storytelling and it makes an other wise dry and dusty subject more alive. 'Visuals' in science can also have a profound effect on getting the message across and is so much more effective than starchy jargon filled journal papers at catching student attention. I agree with Janet above that creating really eye-catching visual representations of scientific processes is highly effective on audiences and makes science so much more accessible.

Story telling and visualisation are very important for science communication as they cast a long lasting and effective impact. Any communication rather is better with these two modes . Communication involves conveying a message and that is beautifully done through visualisation and storytelling.

The scientific rationality is far away frum narrative logic. But it is essential to use a narrative logic by communicating the results of science. There will be no understanding of science without storytelling.

Story telling and visualisation are of crucial importance in medical science and the medical poster is one of the best science communication methods in this field. In a world of, for all intents and purposes invisible enzymes, cells, receptors and bio-molecular structures without the poster and the 'story' it tells it would be difficult and turgid indeed to communicate. I have always found the medical poster to be not only scientifically informative but in many ways as dramatic as any story from either fact or fiction.

Geology and palaeontology: Results not visualized through fotos, maps, drawings, 3D models, movies, diagrams and schemes are hardly percieved at all.

A good technical or non-technical publication is expected to tell the whole story through the figures. (All information is usually given twice: once in the continuous text and once in the figures and figure captions.)

Usually slides of a good presentation include few words apart from headlines (everything necessary is given through the presenter's speech).

Perhaps one negative aspect: Too much focus on visualization can lead to the acceptance and spreading of hardly readable (stylistically poor) writings.

Two facts in brief:

1. The value of visualization largely depends on the reader: some understand better from text, some better from images, some need both.

2. Visualization has dangers: it may simplify or even bias communication of results and it may draw too much attention to fields that lend themselves to easy visualization (e.g. in the case of brain research).

This is taking an interesting direction. As for any communication, the audience determines the form and style of presentation and indeed, there is a danger to 'lie with maps' or any other form of visualization.

Maybe this would be the core message to give in any scientific field. Scientists should be aware of the pitfalls and danger of mixing audiences or not communicating uncertainties. Ultimately, we have to acknowledge that communication is a study field in its own right and requires huge expertise.

Little 'follow-up':

Impressive work on the use of infographics for visualizing environmental data - even without understanding Dutch http://www.pbl.nl/publicaties/2012/nederland-verbeeld. I was told that they (environmental scientists, communication experts and designers) spent an awful lots of time on this.

The big inspiration was: https://www.youtube.com/watch?v=jbkSRLYSojo

Ulf-Dietrich: I don't think the danger you mentioned lies in the visualization itself but in the handling thereof. Whenever and however we communicate, we have to be aware of what it is we're actually saying, and as scientists we need to take even greater care to be precise in our statements. Most scientists simply have more experience in expressing their findings in writing, and even that lacks precision far too often. Whatever you say and however you intend to illustrate it, be it with words or with graphics: never forget to question your doings, be critical and open to critique; you'll become a better communicator and a better scientist.

Hi, visualisation is essential to what we are doing, which is to try to describe the dynamics of learning in open (emergent) systems. We have developed a 3D footprint that tracks 25 variables, simultaneously, across a learning event. It could still do with some automation and interactivity, but its working for us (and for others, too).

See: http://footprints-of-emergence.wikispaces.com/Folder+of+Footprints