Archive for May, 2011

Peak Social

Tuesday, May 17th, 2011

Social is one of those lovely words that can be added to anything to make it better.

Media? Nah. Social media.

Learning? Nah. Social learning.

Networks? Nah. Social networks.

And the list goes on. It’s almost as if social is a condiment to be added to whatever concept lacks spice and flavour.

Google and Facebook are battling for supremacy at the intersection of information and social. Google states its goal as being one of “organizing the worlds information”. Facebook wants to help you “you connect and share with the people in your life”. Facebook is winning – or so it is commonly thought. But it shouldn’t be. And if Google had the conviction to stick to its information worldview, it would win in the long run. Instead, Google has acquiesced its view and adopted Facebook’s.

Dunbar and Shultz argue that significant human evolution in intelligence occurred due to the “computational demands of living in large, complex societies that selected for large brains”. Similarly, Mesoudi, Whiten, & Dunbar’s research states that social information receives preference for cultural transmission.

If this hypothesis holds true, then humanity has gained astounding intelligence benefits because of social complexity.

But.

Humanity has hit Peak Social – the point at which we can gain no new evolutionary or developmental intellectual advantage from social activity.

Perhaps the most fundamental human trait, after fulfilling biological needs of food, shelter, and procreation, is the desire to impose order on and make sense of the world. We have, historically, activated social attributes in order to manage information complexity. Language is social (Wittgenstein and Vygotsky both attribute the value of language in giving birth to thought). Artifacts – paintings, sculpture, and videos – are simultaneously an expression of individual understanding and a means to enact social and participatory sensemaking.

I believe that humanity’s sensemaking is dominant and social is recessive, activated primarily to serve sensemaking and wayfinding goals and activities. My colleague at Athabasca, Jon Dron, prefers the less radical view that the social is a necessary and sufficient condition for sensemaking. Regardless of ones preferred view, there is an obvious relationship between our capacity to make sense of the world and the need for social networks and systems to do so.

Early information overload indicates a departure from social means of learning and sensemaking. Several hundred years ago (if we set aside geographic constraints of language and libraries, it was several thousand years), humanity crossed a threshold where what was known by humanity could no longer be known by a single person. To combat this deficiency, methods and techniques like indexes and encyclopedias were developed. The thing that pointed to another thing had to grow in encompassing scope. An article in Diderot’s encyclopedia came to stand for a book. Sensemaking broke from social boundaries and moved into the domain of non-human devices.

The development of the telegraph, telephone, and eventually the internet amplified and joined communication and information systems. Suddenly, what was communicated was no longer about information only, but was itself information – captured, stored, and analyzable in a database.

And it is here that we hit peak social. We’ve clustered and sub-clustered our social relations. We’ve fragmented our information sources down to tweets and status updates. A tweet now points to a revolution in the Middle East. Or the IMF chief’s actions in New York. While social is lovely, warm, comfortable, human, we need to start thinking about what’s next in human development. While social will be a huge part of it, it must give way to methods that contribute to, rather than cluster, reduce, sub-network, and silo, sensemaking capacity.

Sensors now automatically collect data in the absence of human involvement. Increase in computing power and data quantity promise alternative models of science. Information is no longer something humans seek – it is now starting to seek us.

The next evolutionary surge for humanity will be driven by increased reliance on automated systems for information capture and analysis. The infrastructure – the internet and mobile technologies – is already in place. The data being generated is poorly analyzed by individuals (numerous companies are rather eager to do it, but those pesky privacy laws are a problem).

Why does information flow in networks?

Friday, May 6th, 2011

People like Barry Wellman and Caroline Haythornthwaite have contributed significantly to advancing the analysis of the impact of networks on society. Well before Barabasi, Watts, and Strogatz arrived on the network scene, sociologists (and social psychologists) such as Granovetter, Wellman, and Milgram were developing models to understand how people connect. As a result of this work, terms like “six degrees” and “strong/weak ties” and “networked communities” have become mainstream.

With an understanding of how people are connected we can also gain insight into how information flows through a network. I’m sure you’ve seen analysis of the social networks of board of directors at different companies. Valdis Krebs addresses this in Overlapping Networks:

It is usually beneficial to be connected to those who have a good view of what is going on. Information and knowledge is often shared [intentionally or unintentionally] with trusted others, close by. Information leaks and flows, but never too far. Board members who are connected to other highly-aware Board members, have a higher probability of finding out more — but the range is limited.

Basically, our position in a network, and the overlap with other networks, influences the type of information and people that we can access. This example of the tweets following Bin Laden’s death give a good sense of the structure of information flow.

While networks have always been the backbone structure of society and knowledge, they were situated just underneath the consciousness of most people as the experience of life itself pushed networks to the background. Let me give an example. A farmer in 250 BC would teach his children how to farm (networked learning). The farmer’s family was part of a larger network: religious, agriculture (selling, buying), social, entertainment, and so on. Their position in society was determined by their networked background such as: who they knew, who their parents knew, their connection to community leaders, and their involvement in the army. However, the experience of being part of a network was not fully conscious or even explicit. What mattered was who you knew and your role in society – but the daily experience was likely not an explicit one of “I’m connected by 3 degree to person X”. The key here is the explicit, rather than experiential, encounter of networks.

Today, in contrast, our networks are explicit in tools like Facebook, Twitter, email, and LinkedIn. Most of these services give users the ability to analyze how they are connected to others. We are very aware of how we are connected. Even the act of connection forming requires explicit activity from a person : “Follow X” or “Accept friend request from X”. The online formation of networks is more directive than the offline experience. This morning, while at a local Chamber of Commerce breakfast, I met several people that I’d known by name but had not met before. I followed the social protocol of introduction, shaking hands, and polite conversation. While a connection was made, it wasn’t explicit and didn’t carry with it any lingering sense of connectedness. When I follow someone on Facebook or Twitter, the connection seems more real, more intentional.

The daily reality of being connected naturally raises questions about influence of an individual within a network and how information flows within that system. Klout analyzes influence. SNAPP analyzes the social networks that underpin interaction in a learning management system. Researchers can gain insight into how information flows through a company by email analysis. The prevalence of social network tools and the attention now devoted to analyzing the shape and attributes of those networks – and the evaluation of how information flows – overlooks an important question: Why? Why does information flow as it does? Why does a person decide to share information with her network?

Networks can be analyzed quantitatively to determine connectedness, structural holes/folds, degrees of separation, centrality, small worlds, and so on. I’m interested in the qualitative aspects of information flow. Why did you decide to post on your friend’s Facebook wall? Why did you decide to retweet a resource? Why did members of your network decide to retweet your comment?

What are the qualitative aspects of information objects that determine its likelihood of being shared or amplified within a network?

Let’s consider three elements that are involved in addressing the question of “why does information flow” in a network:
1. The individual. If someone has a large following on Twitter, their message will reach larger numbers of people. However, this is an analysis of how information flows – it flows better when more people hear it. Again, why did the person decide to post the message in the first place? Or, for that matter, how did the person get to have many followers? Artists like Lady Gaga acquire followers simply by fame. They don’t provide much insight into why people have different numbers of followers on Twitter. It’s a spillover of their fame in other spaces.

Let’s look at someone like Alec Couros on Twitter. He has 12000 followers. I have 7400 followers. He has posted over 55000 tweets (wow!). I’ve posted 8300. What are the activities of a person like Alec that give him the higher follower count? i.e. – qualitatively, how does Alec differ from others in his activities on Twitter? Does he have more followers because he posts more often? Because he is talented at engaging with individuals? Is it because he replies to more of his followers than I do and that’s why they continue to follow him? Does he participate in more network sub-clusters (such as the #edtech, #phd, or #learn hashtag communities)? Maybe he’s just a nicer person than I am and people pick that up in his tweets.

Clearly, the activities of an individual plays a role in why information flows…

2. The Context. Context also influences why information spreads. For example, Sohaib Athar live tweeted the raid on Osama bin Laden’s Abbottabad compound. His followers shot up to over 104000 within days. The Bronx Zoos Cobra has over 240000 followers as the voice of the escaped (since captured) Bronx Zoo Cobra. During BP’s Gulf oil spill, the BP Global PR account on Twitter gathered over 170000 followers as it mocked BP. On a far smaller scale, I did a presentation at TEDxNYED (video here) in 2010 and added over 100 followers in one day. TEDxNYED served as a bridge into the K-12 community that I’m not very involved with. I’d love to see an analysis of follower counts in different communities. The K-12 community seems better connected and more active on Twitter than the higher education community. Does a K-12 twitterer have a better chance for quickly building followers than someone in a poorly connected field?

3. The Message. This is really the heart of what I’m trying to understand. What are the qualitative attributes of a message that influence why it is shared. Two attributes come to mind readily:
-Relevance – a tweet about something happening today is more valuable than tweeting that Pearl Harbour was attacked.
-Resonance – this is a complex/fuzzy concept that I haven’t fully wrapped my head around but I know it’s important. When someone posts a link or comment on Twitter, and it resonates with me (fears, interests, beliefs), the prospect of retweeting is increased.

Let’s look at a simple coding scheme of what types of messages people post on Twitter:

a) to express agreement
b) to express outrage
c) humour
d) social grooming (I have an iPad, I met person X today, I went for a run, I ate fruit for breakfast)
e) self-promote
f) raise awareness – general information sharing about topics that might be be relevant for network members

Looking at that list – what would you add?

Suggestion: Let’s create a coding scheme (we can do inter-rater validation if it makes people happy) on why things get posted to and shared on social media (Twitter, FB seem the best candidates).. If we have a coding scheme, we can randomly analyze the posting habits of people on Twitter (i.e. who is completely self-absorbed by self-referential tweets). No doubt, the coding process would be better if it was automated (that way we could evaluate the impact of RTs) – sentiment analysis is a big area of focus for social media firms. Not only are media firms interested in who is talking about GM or BP, but what are the emotions behind posts on Twitter/FB?

Educators are paying attention to social media. The surface level network infatuation won’t generate much value in the long run. Getting at the qualitative aspects of why information flows through networks is a more lucrative direction to consider in transitioning social media use for self and network awareness.

Moving beyond self-directed learning: Network-directed learning

Sunday, May 1st, 2011

We just wrapped up our third offering of the Connectivism and Connective Knowledge Open course: CCK11 (readings, recordings, and archives of the Daily are available on the site).

The course is offered as part of the Certificate in Emerging Technologies for Learning that I developed while at University of Manitoba. Twenty students (the max allowed for the course) enrolled in the for-credit version of CCK11. This means they had assignment requirements including two essays, a concept map, and a final project. The final project has been a requirement of all our previous offerings – and has produced some great resources, including the rather popular Networked Student video by Wendy Drexler – now approaching 100 000 views.

Marking assignments can be rewarding as it requires sustained time and focus. Due to quantity, I usually skim most blog posts and articles posted during a CCK course. Marking, however, requires time and focus, which makes it a good learning experience for me. For example, this presentation – Institutional and/vs Networked Learning – in CCK11 raises an important concern around self-directed learning. Leah makes the statement that the expectations of open courses goes “way beyond self-directed learning”. At first this statement took me back a bit. After all, open courses are require learner autonomy and self-directedness. We want learners to get comfortable with personal wayfinding through complex topics and to utilize tools to splice activity streams in order to fulfill personal learning goals.

After a bit of time thinking about CCK requiring “more than self-directed learning”, it dawned on me that Leah had identified an important distinction. Self-directed learning has a long research and philosophical tradition. Malcolm Knowles figues prominently in discussions, but roots go back to Dewey, and even further, to humanist philosophers.

While connectivism begins with the individual, it stresses the growth of connections and connectedness in learning and knowledge. Self-directed learning explains the attributes of learners who learn at their own pace and interest. Is that sufficient to describe our knowledge needs today? I don’t think so.

When faced with learning in complex environments, what we need is something more like network-directed learning – learning that is shaped, influenced, and directed by how we are connected to others. Instead of sensemaking in isolation, we rely on social, technological, and informational networks to direct our activities.

With MOOCs, we emphasize that early course experiences tend to be overwhelming and chaotic. After all, learners face hundreds of introductions, blogs posts, and reading resources, in addition to dozens of new tools and technologies. As the course progresses, small sub-networks form based on shared interests and goals. Learners also gather in various social spaces that we as facilitators don’t create (Facebook was common in CCK11 as was SecondLife) and in language specific forums – a key requirement with global courses.

To address the information and social complexity of open courses, learners need to be network-directed, not self-directed learners. Social networks serve to filter and amplify important concepts and increase the diversity of views on controversial topics. This transition is far broader than only what we’ve experienced in open courses – the need for netwok-centric learning and knowledge building is foundational in many careers today. For example, the discovery of the corona virus (SARS) was achieved through a global distributed research network. New technologies are increasingly assemblies of innovations that often span millennia – a process that was wonderfully covered by William Rosen in The Most Powerful Idea in the World: A Story of Steam, Industry, and Invention . To be competent, to be creative, to be adaptable, requires that we are connected.

Most importantly network-directed learning is not a “crowd sourcing” concept. Crowd sourcing involves people creating things together. Networks involve connected specialization – namely we are intelligent on our own and we amplify that intelligence when we connect to others. Connectedness – in this light – consists of increasing, not diminishing, the value of the individual.