BayCHI meeting tonight

The BayCHI meeting is happening right now. This is the July monthly BayCHI meeting. The first talk is the Dunbar Number, Unstructured Trust, and Why Groups Don't Scale and the second talk is The Real Nature of the Emerging Attention Economy: Seen As a New Level in the Massively Multiplayer Game Known as Western Culture.

The Dunbar Number, Unstructured Trust and Why Groups Don't Scale
Christopher Allen
- Life with Alacrity blog web site
- Christopher Allen is doing social software design
- What is the Dunbar number?
- named after Dr. Robin Dunbar, there is a strong correlation between size of the brain (neocortex ratio) and the social group size, and found out the mean size of group size is 147
- the Dunbar number has been featured in Malcolm Gladwell's The Tipping Point (which I first heard about from the CASCON blogging workshop conference and was mentioned by Joey De Villa)
- Dunbar number also applies to survival groups like military, mafia and terrorist organizations
- what causes this limit?
- Dunbar speculates that 42% of the time is spent in social grooming
- organization requires social grooming, he is talking about the Sopranos TV show where the mafia spend time together
- this is unstructured trust, it is irrational

Unstructured Trust
- No trust
- Familiar stranger, work at Berkeley, you see familiar people you don't know them but you see them while going to work every day
- as unstructured trust progresses, this causes conversations for discovery with the familiar stranger, this is where you find common interests, you move from familiar stranger to start an acquaintance
- we can now become casual friends
- as casual friends, you are likely to meet other friends, then the size of this group becomes like a band
- this is where the group size increases so much, it becomes difficult to participate in the group
- this is the beginning of hierarchy, a group will find a leader or divide into smaller groups
- band becomes a group, and then a mob, the unstructured trust fails and start to impose structure otherwise becomes chaotic
- what are the limits to group size?
- the Dunbar number gives clues as to when the group changes from unstructured trust to structured trust
- there are limits to group size based on his research on an online game called Ultima Online
- Nic Ducheneaut and his group at PARC have done research on social networks and groups in the world of Warcraft (called PlayOn)
- can do network analysis to find cohesion in groups
- size of groups can have impact on group satisfaction -> this is his own personal hypothesis

Concept: Group Size Problems
- small groups, there are too few people, unable to sustain conversation, you feel alone, there may become groupthink, lack of leadership or commitment
- too many people, becomes too noisy, there is lack of trust, cliques and bad gossip

Concept: Cycle of Flames
- text is missing emotional cues in online conversations
- we over-interpret emotions
- message gets sent, then reply is answer to message that is misinterpreted, and then insults happen
- this leads to a vicious cycle of irony -> sarcasm -> insult -> flame (this happens all the time but people are not aware of it)
- therefore need to be careful of emotional words so don't use emotional words like "should, didn't and forgot"
- when necessary, use smilies to avoid confusion :)

Concept: Appropriate Size
- what software works for small groups which is 2-12 active people and best for 4-9 active people (chat room, teleconference w/Backchannel, cooperative editor, discussion list - flat, blogs - shared only to private, blogs - group authored)
- for medium groups (13-150 active, best for 25-80 active)
- instant message, avatar chat, discussion list - threaded, wiki - single workspace
- for large groups (more than 150 active participants), can't rely on unstructured trust, need to have structure
- discussion list - reputation filtered (ie. slashdot)
- wiki - multiple workspaces
- blog - public (structure is controlled by the owner of the blog)
- social network

This looks very similar with what I'm doing with online communities in my research.

E-mail: ChristopherA AT LifeWithAlacrity DOT com

Q & A:
Q: where does the number 42% spent in social grooming come from?
A: Dunbar looked at the groups of monkeys and extrapolated this out for humans. Social grooming includes getting eye contact, there is this act of feeding each other that helps

Q: Transition from unstructured trust to structured trust
- difficult to say, no easy answers
- look for subgroups in Wikipedia, if look at their social networks, the biggest contributors tend to be cohesive and unstructured size range
- support for unstructured trust involves human contact


Q: What is the role of the inactive (lurking) in the group?
A: 90% that participated in AOL were just watching, as group gets noisy then start to get more lurkers. Everytime he converted a lurker to a participant, got 10 more lurkers. You have to work hard to convert the lurkers.

My Question:
Because it's difficult to create a relationship with people online using text and we use emoticons and animated characters, Are we going to start losing the human language from moving to online environments where we start creating own language (like emoticons in instant messaging), should we be concerned with this? Is this going to become a threat?

His answer:
The English language will continue to evolve so there is no threat is losing the human language. What we should be concerned about is that interaction decreases from the online environments.

Q: How to make lurkers become passive participants and passive participants to become active participants?
A: Give credit to smart questions, try to reward the question by giving a really good answer, this then sometimes converts the lurker. Get the casual participant to become active by appreciating them, giving positive reinforcement.

The Real Nature of the Emerging Attention Economy: Seen As a New Level in the Massively Multiplayer Game Known as Western Culture
Michael H. Goldhaber

- an economy is most generally a massively multiplayer, single-level game that involves some kind of passing of scarce entities between players so as to knit all the players intricately together
- this reminds me of the talk that I went at U of T on virtual worlds and how some of the virtual worlds can become an economy
- there are three types of economies from history: feudal, market-money-industrial, attention-driven economy
- attention is not information going backwards and is not time, attention is not a commodity available for purchase or a resource or something that comes in packets or anything that can be precisely measured
- the attention that matters (from an economic standpoint): it goes from person to person, it involves both mind and body, it "mirrors" motions, emotions and intentions, and attention is retained in memory
- attention is to the whole gestalt of the person
- the attention economy is like a cycle where start with:

attention seeking -> some attention from others -> attention from existing stars (and from some of their fans) -> fans who are attracted by other fans -> attraction from stars' fans

- this reminds me that we really are now in an attention economy because we have so much information and content and bombarded by ads, that our behaviour and actions are based on what we are attentive to
- for example, if I'm attentive to the talk that I'm listening right now, then I will be listening intently and thinking of what the talk means (but I guess I'm being passively attentive, cause I'm writing this blog right now and my attention is on writing!)
- Michael is making a comparison with the various economies and mentioning the forms of wealth. In the attention economy, the wealth is audience sharing
- a measure to compare the levels called TPI (Transactions per lifetime)
- TPI coefficient of economies = transactions per lifetime * pool of people transacted with