December 7th, 2012.
Notes as I scribbled them. Read a proper
review.
**Peter Webster - Digital Resources for Social Scientists at the British Library**
British Library research methods guide under social sciences.
Web about individuals and organisations is fragile, things disappear. Most
site owners ignore requests for permission to archive stuff. New legal deposit
legislation next year will allow scraping the Web and archiving everything
without worrying about infringing copyright. Using this data is restricted;
library premises or print one copy for non-commercial use. Can't use an item
simultaneously with any other user.. (regulation is for print, derp)
Full text search for internet archive... open question, very complicated.
Consultation going on about this. Legislation is very restrictive; need to
look at data derived from dataset and how that can be made available.
Mike Thelwell - Sentiment Analysis for the Social Web
Sentiment is seen as peripheral, and often ignored, but actually it's core.
Emotional reaction to tea or coffee.
sentistrength.wlv.ac.uk SentiStrength - detect strength of +ve and -ve
sentiment in short text. Takes into account that text might not be
gramatically correct. In social media, sentiment expressed in different ways
(eg. emoticons, deliberate misspellings to embed sentiment: haaaapppyyyyyy).
List of +ve and -ve term stems and strengths from -5 to 4. People disagree
about sentiment surprisingly much. Something lots of people tweet about must
arouse sentiment. But for big events, surge of quantity of tweets, but not
surge in +ve or -ve sentiment. Lots of sentiment is implicit.
YT comments are easy to get so good source of social data. No ethical concerns
about getting permission to analyse because it's already public. (hmm, no?)
Longer the text, less well it works. But does have a long text mode, with
slightly different scoring.
Used by Yahoo question answering system to work out best people to give
answers. Companies use it for product reactions.
Would like to see these techniques to smaller scale case studies. The most
focused data stream still has loads of junk like what they had for breakfast..
Why are 1 and -1 neutral instead of 0? Because psychology - two scales, not
one.
Jo Belcher - mixing traditional and digital methods to research hidden carers.
Is online support for parents and carers socially patterned like social media
use? Needs to reach people who don't use the Internet, too, for comparisons.
People respond in the medium they were first contacted. Treat online and
offline results differently for analysis.
Lorenza Antonucci - Using social media in different phases of research process
Digital methods allow you to do something different. A lot of focus on big
data and secondary digital data (twitter). Not about collecting own data, but
using data already available on social media.
Might be a problem fitting secondary data into existing theoretical
frameworks. So need to collect. She's looking at real vs virtual identities.
Not possible to do follow up interviews just over facebook. DM help to select
people. Also cross-national research. Age/generational aspects.
Jeremy Knox - MOOCs
Is learning simply the consumption of information online?
Web enabled sensors. GPS to record where he goes within MOOC - physical
location and digital space together (cool).
Locations tweet when he's in the MOOC.
RFID system that allows office books to tweet content.
Experiments to disrupt and critise. Playful methods to think about the MOOC in
a different way. Assemblage of human, technology and place; learning might be
post-human.
Final outcome of research? Not sure.. a way to provoke thinking in what is
often a closed area.
How is learning affected by physical space?
Sue Thomas - technobiophilia
O'Reilly's topsoil metaphor is cool.
Five categories about how people talk about online experiences.
Lots of nature metaphors. Metaphors aren't deliberate. We bring nature into
computing because we innately want it to be there (biophilia)
Carole Kirk - Digital reflection: A method for arts practice-led research?
Questions and methods come from practical. Tacit knowledge. Capturing creative
processes.
How to leave a trace of an action for reflection?
Digital methods can help.
Not a complete record, but a trace. Digital technologies that involve a high
level of manipulation stimulate greater reflection. Only archives - metadata,
feedback & discussion. Visible record of reflection. Process of creating
records doesn't replace the practice itself. Might trigger embodied memory.
Help to articulate fleeting things.
In arts, practice-led research is more about creating digital data.
Emma Hutchinson - Asynchronous online interviews and image elicitation
Async, like email or forum.
Complement interview with photos, but not used much online yet.
Identity performance of online gamers.
Images help with articulation. Lots to talk about.
Photos that do/don't get uploaded to facebook, why / why not?
Eve Stirling - Facebook profile as research tool
Undergraduate transition to university.
Lots of HE happens on fb. Looking at the every day.
Digital and physical spaces.
Personal fb is not academic, and hidden. Twitter is academic.
Ethnography is about understanding every day culture and developing trust and
rapport with participants.
Fb friends are linked to study. Intrinsically linked. Does becoming fb friend
need a disclaimer? Informed consent? Personal and professional lives blurring.
After study - delete friends?
Me! Digital Media on the Semantic Web.
My slides are here.
danah boyd - Making Sense of Teen Life: Strategies for Capturing Ethnographic Data in a Networked Era
Understanding social networks before there were sns
The rise and fall of myspace.
How much can be made sense of from a distance? Engaging with own friends not
like working with young people. How radically difficult it is to interpret
what she sees. Young people better at encoding the information they make
available, because of adult surveillance. Just because she can see their
content, doesn't mean she knows what's going on.
Observes offline too. Adults help to recruit youths with different
perspectives.
Thought about recruiting online but stopped. Not a good norm to start.
Make sense of online things with offline interviews (ethical things
considered, parents/friends nearby).
Don't begin the conversation with online material; need them to feel
comfortable first; usually an hour into the interview.
How to coordinate data? Serious challenge. Blogging about things as she's
thinking through them; trying to make sense of them in a public way.
Thinking out loud, can be corrected and challenged during sense-making
process. Not just experts, but her participants too.
Controversial piece about shift from myspace to fb. Got picked up overnight
and got over 10k responses. Most people frustrated and angry and didn't
understand where she was coming from. People came forward with quantitative
data that helped. Adults attacked her for being racist; young people responded
with their stories.
How public to make the young people? Don't expose them, no real names unless
their already public figures. Never quotes online material exactly so people
can't search to identify the young people. Visibility has consequences people
don't expect or understand. Can make people more vulnerable by making them
visible.
The young people could choose to make themselves visible through the process
(some do).
Speak for them or help them speak?
Public in the media to make young peoples' voices heard as much as possible.
Never publishes in a closed access journal.
Collaboration is in sense-making, not writing. More in intervention-driven
projects.
Generally don't want to, but a few exceptions - published two papers with
teenagers.
Tension between MS and research?
No, MSR is academic institution. Lots of freedom.
Teenagers expect her to fix xbox.. External perception is more confusing than
internal.
Can make the case for open access in a way that lots of university scholars
can't.
Problems with paraphrasing quotes from websites?
Tries to make quotes more common, found everywhere.
Her ethics about making people not more vulnerable is worth more than skimping
on real quotes. Helps that she doesn't rely entirely on online data.
Says she's not good at articulating her methods.
Ibrar Bhatt - e-focus groups and e-interviews
Separate summer project on student experience for postgraduate research
students, distance learners, part time students globally.
Needed in depth focus group without them being there.
How involved do the students feel in activities in the School of Education?
Different doing focus groups online - affordances and challenges.
Used Adobe Connect. Facilitator echos some questions.
Multidimensional focus group. Participants could also discuss with each other,
and with researcher, over personal chat.
Guidelines - rehearsal, drop-in session, beforehand - recording of everything;
can integrate their video into a transcription.
Temitayo Abinusawa - Social Networking and innovation
Technical background. How organisations use IT to promote activities.
Social networking - Internet was loads of words, chaos. Make sense of the
words, then you can innovate. Can create products and services to meet needs.
Good ideas that need funding. People discuss ideas online. Organisations are
looking for new ideas too. Organisation can search the Web for ideas to create
innovation.
Outcomes: organisations can create more for less.
Feedback is consumer interaction that takes place on the Internet. Dell
IdeaStorm - turn your ideas into reality.
Openness is important, not exploitation. (Seems to be reward focussed, rather
than transparent exchange culture, query).
People don't read t&c so think they're being exploited.
Evelyn McElhinney - Social virtual worlds: a new place for the avatar researcher
Focus groups in Second Life with avatars collecting, sat in chairs like IRL.
Most people aren't roleplaying, their avatars are just themselves.
Closing Discussion
Ethics
Commercial research is sometimes ahead with digital methods practices. Things
happening one end aren't being noticed by other end. Need to communicate and
look out for each other. Mixed methods research.
Digital methods need legitamising to be taken seriously. Sometimes people not
knowing much about it can help to get stuff done.