Showing posts with label data. Show all posts
Showing posts with label data. Show all posts

Wednesday, 16 October 2013

Standing on the shoulders of seagulls


I'm pleased to present the first guest post on this blog. Here, my friend and colleague (and Natural Sciences student) John Tweedie shares his thoughts on the challenges of communicating science: both within the scientific community and to wider audiences. John also blogs on science, Scottish football, birdwatching, and anything else that takes his fancy.



Knowledge is next to useless if it’s not shared. Knowledge is created by sharing and standing on the shoulders of the proverbial giants who came before. From town criers proclaiming a monarch’s decisions and clans and tribes sharing knowledge through song and poems to the advent of the printing press and the development of the world wide web, people have a thirst for knowledge and an urge to share it with others, be they friends or opponents, colleagues or competitors.

Emerging from the shadows of alchemy through the works of Aristotle and Archimedes and onwards through the Enlightenment to the modern day, science has developed an extensive body of knowledge and various means to communicate, both within science and to the general public.

The scientific method is the accepted way in which science is conducted — a question is posed or an observation is made and a scientist will then develop a hypothesis to find general patterns that underlie the question. Through experiment and further observation, they will refine their hypotheses and so become more confident in their outcomes.

A major part of this process — indeed the main way in which to further develop the hypotheses and build towards the more robust ‘theory’ — is communicating the findings.

There are several ways this can be done. One is to take part in conferences where peers can ask questions about methods used when conducting an investigation and their outcomes. Often these presentations can be used to test early parts of work — is the work well received, or are people picking lots of holes in it? The feedback can send a scientist off in different directions — perhaps their methods have been flawed or their statistics do not support their work. Further revisions and presentations can lead to the work being readied for publication for the wider scientific community to study.

The general process of getting a paper published is regulated by the major journals and through peer review. A paper will be sent out to reviewers who will scrutinise the paper and decide whether it’s fit for publication.

Whether another scientist can or cannot replicate the findings will lend weight to the hypothesis or result in refinements of the hypothesis. Later studies replicating the original investigation may include errors in replicating the experiments, in which case these findings must be published too. In science, negative results are every bit as important as positive results. A good example of this was the recent search for the Higgs boson — a negative result would have meant scientists having to revise their hypotheses and models, which in itself is a continuation of the process of building knowledge.

The most important part of a paper is not the conclusions, but the methods. This is where a scientist demonstrates how they went about answering their question. It’s this part of the paper that allows other scientists to attempt to replicate the experiments and observations conducted. It’s important to note that most scientists work in teams and collaborate within their institutions and with colleagues across the globe, and that published papers often represent the work of many people.

There’s a tension between new discoveries and building on previous knowledge. Scientists must be familiar with the current thinking and all that has gone before in their field.

John James Audubon is well-known for his work in the early 19th century as a field ornithologist and artist. To test whether the turkey vulture had a good sense of smell he put out hidden carcasses, and concluded when the birds didn’t find the meat that they did not have a well-developed sense of smell. Unfortunately for him, his experiment was flawed, because if he had known more about these birds he would know that they prefer fresh carcasses rather than putrefying flesh, and it turned out that the birds he was observing were in fact another species which have a poorer sense of smell than the turkey vulture.  He was also under the mistaken belief that animals can only have one well-developed sense; in other words, if a bird has a good sense of sight they cannot have a good sense of smell.

Much science is highly specialised and only of interest to other scientists, but enough is of potential interest to the general public or considered ground-breaking or to have real-world applications that could solve urgent problems facing humanity and the world today. These papers are picked up and summarised first by popular science magazines which will present as much of the science as possible but for a non-specialist but highly interested and educated audience.

Next up are newspapers and the media — they’ll often simplify the findings, bringing out the practical applications of the science or demonstrating in what way the findings may or may not have revolutionised our knowledge about a particular topic. They’ll often put the science in context with what came before and what the science could lead to in the future.

This is where science communication to the public becomes really important. Science thrives on using precise and specialised language, often underpinned by difficult mathematics or conceptual ideas. The News at Ten is not going to devote time to talking about the intricacies of quantum mechanics: to properly understand such a topic a person really does need an education covering the basics — in this case physics — what an atom is, what it is composed of and how they interact with each other and the transfers of energy.

I have such a lot of respect for many of the science correspondents who can convey complex ideas in the few minutes they get, usually at the tail-end of news programmes. What they do is explain what the science means through using analogies or how the findings are going to find practical uses in the future. Are the results of the findings going to result in faster computers and better communications technologies, are they going to enable effective medical treatments to be created? Most people’s lives have been enriched by the work of scientists: indeed many people are still living due to research, so this is often where the excitement of ‘big science’ is really conveyed.

As much as the general media help to shape public interest in science they also have a responsibility to present it accurately. It’s all too easy to oversimplify the science to make it seem as if a study is something that has a ‘common sense’ answer, something the social sciences are particularly vulnerable to. They could misrepresent the science, leading to confusion and misunderstanding, perhaps undermining public confidence in scientists.

The media can also suggest that ideas are contentious and that there are dissenting voices out there when really there is none within science itself. This is common with climate science where the media, in a mistaken sense of balance and fair play, will allow sceptics or deniers lots of airtime, potentially giving the impression that the scientific consensus is undecided. The BBC has come under particular criticism for this, highlighted in a report by Steve Jones. Many of the sceptics or deniers are not scientists, but are opposed to the idea of what the science represents, either through ideological, political or economic beliefs, and they are particularly concerned with how countries will respond politically or economically to putting in mechanisms to slow down human-caused global warming.

Andrew Wakefield’s now discredited work on the link between the MMR vaccine and autism is a prime example of how irresponsible communication can lead to health scares, and ultimately to deaths. Parents honestly think they're behaving in their children’s' best interests, and often talk about making informed choices for their children, but when the findings of science is misrepresented it can have dire consequences. Many people do not have the scientific education to make truly informed decisions on matters like these and so the media has a responsibility to present findings accurately. This is an example where the peer review process had failed: it should have never have been published in the first case.

As such, I’d like to see scientific communication be a part of all science undergraduate training. Many scientists like to distance themselves from the media and the general public, enclosing themselves in their labs and ivory towers and working in the pursuit of pure science. However, with much science being publicly funded and ultimately having practical applications, scientists should be able to communicate to the press and media. Not everyone can be a Richard Feynman or Carl Sagan, but the ability to communicate clearly to a wide audience is something that should be nurtured. Science has many ideas and words that are used in a specialised way — theory and uncertainty are two examples which the general public do not understand — but scientists should be able to clearly explain what they mean in the context of their work. Often critics say things like ‘it's only a theory’ to shed doubt on evolution, or the public take uncertainty to mean that scientists aren’t sure about their results. These ideas and indeed all science must be clearly communicated.

There are many other mechanisms of science communication. These include popular science books, sometimes written by journalists or historians of science who bring together whole fields and present the science in context, but often by working scientists themselves. Textbooks find their place in university libraries and are only really read by students and other scientists.

Radio is where the voice of scientists is often heard by the general public, rather than their work being interpreted by correspondents as on TV. It’s this medium where they get more time to explain their ideas to the general public, and it’s here that the really good communicators excel.

Some excellent communicators and working scientists, such as Brian Cox and Alice Roberts, are able to work across all media and they really engage the public, often enthusing about science and encouraging the next generation of scientists. They present big ideas that are exciting — normally little knowledge is expected of the audience, although sometimes, like Jim Al-Khalili, they’ll present more difficult concepts, expecting the audience to keep up.

All of these examples show just how wide and varied science communication is. Science is incredibly important to our societies — its developments enable our way of life to be possible — we live longer, we can access information on any topic at any time in just a few seconds, we can travel hundreds of kilometres in a few hours, we can talk to our friends and families instantly living on the opposite side of the planet, and in the developed world we have more leisure time due to developments in technology, often derived from pure science.

We’ve put men on the Moon and to the bottom of the oceans, we’ve sent probes to other planets, one has just left the Solar System, and a rover is sending back information from another planet. We know how old our Solar System is and how long it’s expected to last.

All of this science has had an amazing impact and it will continue for as long as we keep asking questions and enquiring about how the world and the Universe works, with communication being at the heart of the process.


I suppose I ought to try and offer some sort of partial defence of the media's coverage of climate change. It's very hard for those of us on the other side of CP Snow's two cultures to accept that contentious issues can be genuinely settled one way or another, and we're right not to accept it in other areas: we journalists have a healthy scepticism towards authority which helps keep politicians and bureaucrats in check the rest of the time (or, at least, it does until they decide it won't any more).

There clearly is still a genuine debate about what to do about climate change, even if there's no longer a debate about the science itself (or, at least, the science so far rather than its extrapolations in the decades to come). I hope this secondary debate — essentially, should we risk short-term economic growth by trying to stem carbon emissions, or should we accept a warming world and instead spend our wealth on better adapting to it? — will now become the focus of the mainstream media's coverage. This Economist leader, for instance, seems to get the balance right.

Friday, 26 July 2013

The open society and its enemies

Thrill-a-minute hedonist that I am, I've spent the week reading Facts are Sacred, the Guardian's new book on data journalism, and The Signal and the Noise by clever clogs Nate Silver. The usual low standards of Guardian subbing are met, but ignoring that there's some interesting content. I note that the New Statesman, New Scientist and several marketing magazines have also devoted recent front covers to big data.

Nate Silver is impressive on the limitations of data; what he really advocates is an approach combining both quantitative and qualitative analysis, as well as a Bayesian method of improving one's probabilistic forecasts as the evidence changes. His advice — again, equally applicable to anyone in business, politics, communications or journalism — is to be consciously aware of your own biases if you want to overcome them and see the situation objectively. You hardly need me to add to the mountains of textbooks already written on how institutional groupthink and inertia have brought down many once-mighty companies. The challenge is to constantly apply this scepticism to your own positions.

At the risk of coming over all Blairite, there is also an open/closed distinction at work here, and the point applies as much to mindsets as datasets — and it cuts completely across traditional left/right or even liberal/conservative lines. As a recent Economist piece put it: "Our ideological sympathies are not good predictors at this point of how we feel about issues of digital privacy and electronic freedom. The fact that these issues don't have a clear ideological colouration yet is important because they are among the most crucial issues of the 21st century."

I would summarise this as the division between those whose instinct is to make information more widely available, and those whose instinct is that it should remain the preserve of responsible gatekeepers of knowledge — between:


We could list more examples, but the point is that this battle against banal illiberalism will define the era of big data and big brothers. Those of us working in communications are at the frontline, so we might as well get ourselves on the right side. If you're a journalist: understand how data can be used both to find and to illustrate stories, carefully craft and submit FOIs,1 resist the lure of churnalism; if you're in PR or marketing, convince your bosses to make openness and clarity an integral part of your brand, develop a responsive social media presence that makes this a reality, and never let the cover-up eclipse the crime.

As YouGov's Peter Kellner says: "knowledge is not only preferable to ignorance in principle; it also makes for a healthier society in practice". Encouragingly, one of the 10 design principles on the new gov.uk site is: "Make things open: it makes things better." What we really need is for this message to filter up to ministers and senior civil servants, and down to middle managers.

The point isn't that we need to be anti-capitalist or anti-government in principle; just that, to make both of these essential parts of our society work better, we need to keep a close eye on them from inside and out. As citizens we have to hold public bodies and global companies to account every bit as ruthlessly as we hold producers to account when we choose whether or not to buy their goods and services. As employees we need to fight institutional cultures that prioritise the organisation above the consumer or the public.

Erring on the side of this openness is good for society; whether it's good for individuals is an entirely different question.

1 I recently did this in my capacity as an elected member of Dowanhill, Hyndland and Kelvinside Community Council, and after waiting nearly a month was told by Glasgow City Council I'd need to pay a fee of £90 to get the answer I was looking for. But don't let that put you off. ;-)

Monday, 28 January 2013

Save the data

This week we learned that Glasgow (my home town) has been volunteered to host a pilot project to collect and use urban data more efficiently. The idea is to link up real-time info about transport, traffic and energy to make delivery of related public services more responsive. Even in a medium-sized city like ours, the amount of relevant data my fellow citizens generate would seem to make this a huge job.

According to some data I just found on the internet, 90% of all the data that exists was created in the last two years. The challenge of sorting through all of this stuff, and condensing or consuming it for some useful purpose — for research, for commerce or just for fun — is becoming one of the great communications challenges of the age. Essentially, the job of any professional communicator has always been to sort through a lot of data and transform it into useful information: in journalism this means bringing out the salient facts of a story, in marketing the key selling points of a product or service, in PR a bit of both. But how do we go about sorting through the deluge of data that exists now?
  
JISC's rather good newsletter has some ideas about how we as individuals and institutions might go about this daunting task. I liked point 6, though it doesn't seem a particularly risky claim to suggest the visualisation of information will increase. The project to display interactive information about every bomb dropped on London in WWII, explored on the next page of the newsletter, is a fascinating example.

It's worth remembering that not all of this is new. This Economist article from 2007 on the history of graphic displays of information contains some examples that are just as impressive and eye-catching now as when they first appeared centuries ago. It's hard to think how today we could better illustrate the disaster that was Napoleon's campaign in Russia than by Minard's chart, while Playfair's beautiful depiction of wheat and labour prices seems to anticipate the Manhattan skyline. The dismal science can be fun too.

The JISC piece also leaves out the problem of institutional bottlenecks: right now I'm trying to explore visual.ly and easel.ly, but the browser on my work computer is obsolete and my attempts to get IT to upgrade it have so far proved futile. The potential for big data is exciting, but any big organisation generating the stuff is also liable to be sluggish in working out how to deal with it.  Big is not always beautiful, as this Guardian article explains. Mountains of data are of little use if your local council only consults 42 people about a major planning decision, then ignores them and does what it wants anyway.

One final point: as with all technology, the information revolution merely magnifies existing trends in society and human nature to produce new opportunities and new dangers. Might the disturbing extent to which American politics has become polarised be at least partly a feature of the ease with which readers, viewers and listeners can now avoid accessing opinions they find unfavourable?