Thursday 23 February 2017

Programming as Social Science - new methods network

Phillip Brooker is a Research Associate at the University of Bath working in social media analytics, with a particular interest in the exploration of research methodologies to support the emerging field. His background is in sociology, drawing especially on ethnomethodology and conversation analysis, science and technology studies, computer-supported cooperative work and human-computer interaction. Phillip has previously contributed to the development of Chorus (www.chorusanalytics.co.uk), a Twitter data collection and visualisation suite. He currently works on CuRAtOR (Challenging online feaR And OtheRing), and interdisciplinary project focusing on how "cultures of fear" are propagated through online "othering".

Digital data and computational methods are increasingly becoming consolidated as essential elements of social science research and teaching. However, the algorithmic processes through which digital data are extracted, processed and visualised are often ‘black boxed’ and obscured from researchers who use those tools, which hinders our understanding of how they might be handled methodologically. Hence, there is an already-high and ever-increasing need for social scientists to engage with computational tools as a “critical technical practice” (Agre, 1997). In other words, since we are now pretty much completely reliant on software as part of our everyday research and teaching practices, it is all the more important that we were able to unpick and interrogate how these software packages operate, in order to better account for our data and research practices!

To this end, myself and Jonathan Gray (both at the University of Bath) have set up a mailing list/network called “Programming as Social Science (PaSS)”, for researchers interested in software programming both as an object of study and as a tool that we can learn and use within social science research. Here, we’re capitalising on lots of good work that has already been done in fields such as Science and Technology Studies, New Media Studies, Social Media Analytics, Software Studies, Ethnomethodology, Human-Computer Interaction, Computer-Supported Cooperative Work, and so on. All of these fields (and many more we haven’t listed!) have contributions to make in regard to understanding how we might critically leverage programming skills as part of social science teaching and research. So the PaSS mailing list/network has been established to act as a (low-traffic) hub for discussing these kinds of ideas, as well as sharing resources, updates, announcements and initiatives around programming in the context of social research.

If you’d like to join in, you can sign up via the following link: www.jiscmail.ac.uk/PaSS. Please feel free to invite anyone and share widely; the computer geek in me is very much looking forward to chatting about programming as part of my work!

Thursday 16 February 2017

Visualising Facebook

Daniel Miller is Professor of Anthropology at University College London. Recent books include Social Media in an English Village (UCL Press 2016). Miller. et. al. How the World Changed Social Media (UCL Press 2016). With J. Sinanan Webcam (Polity 2014) Ed. With H. Horst, Digital Anthropology (Bloomsbury 2012). With M. Madianou Migration and New Media (Routledge 2012) Consumption and its Consequences (Polity 2012), with S. Woodward Blue Jeans (California 2012) Tales from Facebook (Polity 2011). He recently completed a volume about media in the social lives of patients with a terminal diagnosis, forthcoming as, The Comfort of People (Polity 2017). @DannyAnth

This March will see the publication of a new book called Visualising Facebook, which I have written with Jolynna Sinanan. It will be available as a free download from UCL Press. One of the key arguments from the larger Why We Post project, of which this book is one out of eleven volumes, is that human communication has fundamentally changed. Where previously it consisted almost entirely of either oral or textual forms, today, thanks to social media, it is equally visual. Think literally of Snapchat. So, it is a pity that when you look at the journals and most of the books about social media, they often contain either no, or precious few, actual visual illustrations from social media itself. One of the joys of digital publication is that it is possible to reproduce hundreds of images. So, our book is stuffed to the gills with photographs and memes taken directly from Facebook, which is, after all, our evidence.
For example, as academics, we might suggest that the way women respond to becoming new mothers in Trinidad, is entirely different from what you would find in England. In the book, we can reproduce examples from hundreds of cases, where it is apparent that when an English woman becomes a mother she, in effect, replaces herself on Facebook with images of her new infant. Indeed, these often become her own profile picture for quite some time. By contrast, one can see postings by new mothers in Trinidad, where they are clearly trying to show that they still look young and sexy or glamorous, precisely because they do not want people to feel that these attributes have been lost, merely because they are now new mothers.

In writing this book we examined over 20,000 images. These provide the evidence for many generalisations, such as that Trinidadians seem to care a good deal about what they are wearing when they post images of themselves on Facebook. While, by and large, English people do not. But this becomes much clearer when you can see the actual images themselves. Or we might suggest that English people are given to self-deprecating humour, while Trinidadians are not. Or that in England gender may create a highly repetitive association between males and generic beer, as against women with generic wine. In every case, you can now see exactly what we mean. We also have a long discussion about the importance of memes and why we call them `the moral police of the Internet’. How memes help to establish what people regard as good and bad values. This makes much more sense when you are examining typical memes with that question in your head.


To conclude, given the sheer proportion of social media posting that now consists of visual images, it would seem a real pity to look this gift horse in the mouth. Firstly, it has now become really quite simple to look at tens of thousands of such images in order to come to scholarly conclusions. But equally, it is now much easier to also include hundreds of such images in your publications to help readers have a much better sense of what exactly those conclusions mean and whether they agree with them.

Friday 3 February 2017

Mine your Data – Why understanding online health communities matters

Originally posted on the NatCen blogsite on 10/11/16 
Aude Bicquelet is Research Director in the Health team. Prior to joining NatCen in 2016, she held a fellowship at the LSE (Department of Methodology) where she taught courses on Research Design, Mixed-Methods and Text Mining approaches. 
Aude specialises in the analysis of ‘Big Qualitative Data’ on health related issues and has worked with professional and regulatory health bodies such as the National Institute for Health and Care Excellence (NICE) and the Royal College of Physicians.  Methodological and substantive outputs of her research have been published in academic journals; she has also published a book on ‘Textual Analysis’ with Sage.
In addition to her interest in Health policies she is interested in Social and Political attitudes and has researched widely in the areas of political participation and élites’ attitudes towards the EU. 

A staggering 73% of adults in the UK turn to the internet when experiencing health problems. Whether it is to check symptoms, find out about available treatments or share experiences about living with a particular condition, the internet has become the first port of call with many turning to the web before they even consider going to see a doctor. While many of these conversations take place on health-related websites such as Patient or Netdoctor, people suffering from health conditions also share their experiences on social media – and health practitioners should take note.  
Earlier this week I presented findings from a recent study looking into how people use social media to discuss health issues at the ESRC Festival of Social Science. In this study, funded by the NCRM, we used text mining techniques to analyse comments about chronic pain posted under YouTube videos.  
We found that chronic pain sufferers use YouTube to describe their experiences and vent their frustration. We analysed over 700 YouTube comments, and found they can be sorted into one of five categories:
  • Sharing Experiences: commenters thank each other for sharing their experiences in the videos posted on the website, emphasising tolerance and empathy for chronic pain sufferers.
  • Expressing Frustration: chronic pain sufferers expressed their frustration in their own words. These illustrate how YouTube and other social media offer new avenues for communicating pain outside clinical contexts.
  • Coping with Pain: chronic pain sufferers used social media to share their daily practices to cope with chronic pain.
  • Alternative Therapy: commenters spoke openly about their use of alternative medicines, illegal drugs or alcohol to manage their pain. The often conflicting relationship with clinicians – who were perceived as over- or under-medicating – was also common in this category.
  • Risks and Concerns: they also discuss the risks associated with different types of medication – in particular, addiction and overdose - along with increased risks of depression associated with some treatments against pain.
The insights gained from social media research provide important substantive information for health practitioners. People communicate online in a way they don’t during interviews with researchers or during doctors’ appointments. Online forums and social media are rife with information that’s difficult to obtain through traditional research techniques where social desirability, fear of judgement or stigma, and wanting to be seen as ‘functioning well’ may influence what people are willing to say.  From a purely practical perspective, they also provide freely available naturally occurring data with access to (at times) to hard to reach groups.
Of course, there is a great deal of uncertainty around how to harness the opportunities of analysing the wealth of health information posted online in a representative, robust and ethical way.
Despite their usefulness and efficiency, analyses of Internet comments on health forums do raise a host of concerns such as representativeness – where the views of one cohort in a population having access, technical skills and inclination to post comments on Internet websites are over represented while the views of others are excluded (i.e. the so-called ‘digital divide’) and consent – where, online commentators may not expect to be research subjects.
Nevertheless, the explosion of Big Data and the popularity of online communities might precipitate the need to integrate social media analysis and health research in the near future. For instance, it has been shown that patients who visit their doctors with inappropriate or misinterpreted information from the internet will do little to enhance doctor–patient communication (see Ziebland 2004). But, doctor-patient communication could be improved simply if health professional themselves were better informed about the common fears and sometimes the common ‘myths’ disseminated on online health communities.

Watch Aude’s presentation from NatCen’s event ‘What Social Media Can Tell us about Society’, live from Twitter’s London HQ. This event was part of the ESRC Festival of Social Science
If you’re interested in how social media research can help you, please get in touch: aude.bicquelet@natcen.ac.uk or new-business@natcen.ac.uk