Online Group Chat Experiment (v. 1)

Posted on Tuesday, April 14th, 2009 at 10:41 pm by Jeremy in process, prototype | 2 Comments

This past weekend we conducted our first experiment to better understand the behavior of people participating in an online group chat (using an instant messaging application). Two groups of 4-5 participants were recruited and asked the same series of three questions, in each case asked in a slightly different order. The discussions for each question lasted approximately 10 minutes before being abruptly halted by the moderator. While this batch of participants was gathered from the SMSC team’s personal network of contacts, all of the participants in each discussion had never met and their identities were generally unknown to each other.

The questions we posed were:

  • If newspapers continue to go out of business, what will replace them?
  • Are zoos a humane institution?
  • From the perspective of the United States, was it the right decision to drop the atomic bomb on Japan?

The newspaper question was chosen due to its timeliness; the issue has been widely reported in the media recently with the shuttering of several print newspapers. The second question was chosen because of its ethical dimension and the fact that it may not be a topic that many have given sustained consideration to. Finally, the third question was chosen for its potential “polarizing” effect. The order that this question was asked switched from the third position to the first position in the second group chat.

We’re still digesting the results of this first iteration, so I’ll refrain from drawing any conclusions at this point. Overall, the experiment was highly instructive in establishing a baseline for our future work and we’ve already begun to plan for successive iterations. It’s clear that we need a different strategy for recruiting a more diverse group of participants, and the importance of the polarizing question is increasingly apparent given our attempts to break down civil discourse within social media in order that we may rebuild it in a different fashion.

Earlier in this process we had been very excited about a game-play strategy for structuring mediated communication between strangers. I think this approach is still potentially valuable, but we first had to make a few smaller steps in order to better understand the terms (and the problem); the game had too many variables for us to contend with at this stage. Stay tuned for future iterations in sited in public that maybe reintroduce game play.

Posted below is a selection of short video documentation from the experiment.


SMSC – Discussion in preparation for an online group chat experiment: the importance of asking a polarizing question.


SMSC – Discussion in preparation for an online group chat experiment: formulating the polarizing question.


SMSC – Online group chat experiment

2 Responses to “Online Group Chat Experiment (v. 1)”

  1. # On April 17th, 2009 at 6:52 pm Alie wrote:

    You guys have to stop giving me caffeine.

    In other, more pertinent news, I am really happy with the first experiment. I think the two very different results we witnessed are really suggestive of what some of the underlying issues are in on-line debate. To be specific: the participants and their level of education as well as their engagement with the topics under discussion; how the questions are framed; and the devices and language people use specifically in the context of on-line discussions. For me this is a great start to evaluating how important each of these and other factors are in how fruitful, diverse, and respectful on-line dialogs can be. So! I’m excited for more testing and for the results to our surveys.

    Moments of particular interest to me during the discussions were:

    The sharing of links and citing outside sources made it easy for people to back up their points. Most people don’t carry back issues of the NY Times in their pockets so this is unique to on-line dialogs (I wonder if the other participants viewed these during or after the discussion or even at all.)

    During the second discussion, one of the participants got all meta on us and attempted to lay down rules for the discussion, that each participants would speak their peace and then “pass.” The rest of the group decided the free for all was best.

    Attempts to maintain contact between some of the participants in the first discussion through facebook. (easier than exchanging numbers?)

    Readiness/reluctance to answer questions directly/immediately. There was a whole lot of questioning the question.

    So, yeah, excited to see more and see if any patterns develop!

  2. # On April 23rd, 2009 at 2:24 pm Ona Krass wrote:

    Once again, I find it fascinating that in SMSC trying to develop a social media project about democracy in a public space or in terms of debate, we ourselves exercise the Social Rules that we are expecting or hoping for in our experiments. The first videos that display a table of minds discussing how to carry out a group project, shows a commons of ideas. OH! The “Meta-narratives.”

    I think our first test was a huge success in terms of interest and was an imperative step, but at the same time I think it helped us define where now to move and what to move away from. I think Sunday, if our participants show, we might possibly get some controversy. I’m hoping for! Mainly, in the sense that it will require more engagement and spice up a relationship between our experiment and the reality of our problem.

    So Onward!