Social Media and Society Conference – Day 1 Summary

The first session I attended, on NodeXL, let by @marc_smith and @alexfenton, was excellent. I have a three month product key for the Pro version because I attended the session – almost worth the price of admission.

I understand the associated product requires promoting, but Marc Smith did an excellent job of contextualising the importance of work on social media conversation and how we view crowds, drawing on social network theory.

I brushed up on my rusty sketchnoting skills to bring you this TL; DR version:

Sketchnote of NodeXL #SMSociety session by Anna Bartosik

Connecting with the content: What I got out of it

Looking at an image of a crowd tells the observer one story, and we can deduce some information (i.e. signs, logos on shirts, age of participants in crowd, emotions, etc.):

Group of people attending concert by chuttersnap on Unsplash

…but the still photo can’t tell us:

• the political views of people in this crowd
• which people know each other
• which communities of people we see
• if the communities communicate with each other
• if communities or people communicate ABOUT each other
• where the crowd comes from, geographically
• who sits at the centre of the conversation (not the most popular)
• if communities begin to form because of people who have “betweenness”
• who is watching but isn’t participating

The QUAN part of the session was centred arount Network Theory but from a sociological perspective. Marc Simon talked about network theory research and how it has been traditionally viewed as a quantitative pursuit because of the sheer volume of data. A network view shows us the shape of connections around us, and, instead of just looking at numbers, he asked “What does a hashtag look like?”

The NodeXL Graph Template Gallery provides a menu of recipes for different visualisations of hashtag data. NodeXL, like other data mining programs, provide a visual representation of annotated connections that show connections among a crowd, between people in a crowd, and it can also measure “birds of a feather” and between-centred individuals, who are responsible for making these bridges to other groups. If they are removed from the graph, the groups become silos.

There are six different kinds of SM networks on the NodeXL Gallery, the recipes for which can be imported into NodeXL Pro for analysis. In addition, NodeXL also groups people with three different preset lists of words, which can be edited, and additional lists can be added. This would be useful when researching a particular topic, like a political movement; lots of those these days, and the data is overwhelming. This type of software (competitors include Netlytic and Gephy) helps make reading this data a possible task for humans.

Twitter is the easiest to analyse; Facebook analysis can only happen with pages, and that emphasises silos, not connections or individuals, because of what Facebook allows and doesn’t allow. There isn’t a way to inspect Instagram in the same way as Twitter because the line for what is viewable/not keeps changing.

Alex Fenton talked about ethics and his research with Twitter and participants – research ethics boards say anonymise participants; participants say don’t anonymise me – I want to be included and I was public on Twitter for a reason. I’ll address this in my upcoming ethics post.

Footprints in the sand by chuttersnap on Unsplash

Descriptive writing

Another way for me to contextualise how social content mapping and the analysis process works is to connect it with my area(s) of expertise (because once you’ve studied instructional design and cognitive processing, you can’t not try to make connections with what you already know).

I used to walk English for Academic Purposes (EAP) students through a descriptive writing activity to practice free writing. I would use prompts to help inspire ideas and directions for writing. The questions associated with Social Network Analysis (SNA) are similar to trying to create a story from a framed frozen image.

In the sand, Who is at the beach? What are the ages of the people there? Are they related? How do you know? Which set of footprints has the most connectiveness – who do the other footprints flock to? Which direction are most of the footprints heading? Do any of the footprints retrace their steps? Why do you think they do that? Were the people running or walking? How do you know? Is one of the individuals taller or heavier than the others? What clues do you have to answer this question? Are there any couples walking together? Are there any footprints which don’t cross paths? When were these prints made?

This different perspective on an image of a crowd is without faces, but still tells a story, an interesting one. These footprints show connectivity and how networks are formed. And much like Twitter and conducting research, there is not much time to capture the moment because the footprints will be washed away, and data from Twitter after 8-9 days is not available, and trying to access it is expensive.

(sentence) patterns

Gathering insight into the nature of the conversation in these networks and understanding what SNA does takes me back to my TESL pedagogical grammar class and sentence patterns (I believe June 10, 1996 – the day I learned this – was the day I decided any examples I wrote on the board would be interesting):

How I spent my summer, 1996 by Anna Bartosik

Who (N1) –> did what (IV)–> with whom (N2)? When? This is is what we want to know.

There are three resolutions when using NodeXL and other social mapping tools: the person, the network, and the graph. We can also see two different people talk about the same thing in two different ways, but not with each other. How? By examining hashtags.

My next post from the conference will focus on apps and permissions and tracking.

Advertisement