Collective Intelligence

Collective Intelligence and the Wisdom of the Crowd

The idea that the participatory web can tap into the knowledge of the group is appealing. However, while this is a section on justification, what follows is not universally positive. We examine what some recent authors have to say and relate some of these ideas to well established issues. There are challenges, but there may also be solutions.

Is there an efficient way for all of us to aggregate of what each of us knows? One argument for participatory web experiences is that participatory experiences bring us into contact with what others know. In theory, the ability to easily tap the unique knowledge many individuals collectively hold offers advantages over what even experts know.

The "Who wants to be a millionaire" example

Anyone who has watched the television program "Who wants to be a millionaire?" has observed the benefits of collective knowledge. As you probably know, the program allows contestants to answer a series of progressively more difficult multiple choice questions and if the contestant can answer 15 questions without an error the prize is $1,000,000. To spice things up a bit, the game offers a few optional strategies. Once during the game, a contestant can randomly remove two of the wrong answers, call a friend for assistance, and poll the audience. What is of interest here is the comparative benefit of asking a friend and polling the audience. It is assumed that contestants will identify experts they know (probably the friends they feel are most intelligent) for their phone calls and that these individuals would be more knowledgeable than the typical member of the audience. What is easily observable from watching the program is that polling the audience is generally the most successful "life line"? The interesting question is how can a bunch of average people more successfully identify a correct answer than a single smart person?

Surowiecki's Wisdom of Crowds

This question is one of the topics addressed in a recent book by James Surowiecki - ''The Wisdom of Crowds''. The author argues that the knowledge in a group can be aggregated under certain conditions. These conditions include:

These conditions can be applied to the "Who wants to be a millionaire" setting. It can be assumed that the audience members come from a wide variety of backgrounds and enter their votes independently. The display tallying the vote totals for each response aggregates the information for the contestant.

It may seem promising that the Internet and social networking opportunities made available through the Internet offer productive ways for individuals to pool what they know in order to function at a level of understanding not inherent in individual participants. Perhaps there are ways to harness the Wisdom of the Crowd.

Crowds do not always exhibit wisdom

As you might expect and probably sense intuitively, life is not this easy. Some of the complexities and challenges of integrating knowledge is explored in another recent popular book - Cass Sunstein's Infotopia (2006).

Before I outline some of Sunstein's arguments, allow me to again see if I can tap into your personal experiences to identify some of the challenges in capitalizing on the knowledge present in a group. I assume you have some experience with the CNN and Fox news networks. If you share my bias, I am also guessing you might predict that those with a Republican political bias prefer one of these stations and those with a Democratic bias prefer the other.

An important question in a free choice information environment might be - Do individuals made choices to expand their personal perspectives or do they make choices that shore up their existing beliefs? Do individuals seek information that makes their understanding more nuanced or more extreme?

Sustein offers a mixed review of online knowledge aggregation and makes a convincing case that in many cases "deliberation" (interaction among participants) makes us stupid (my term). For example, in reference to blogging he uses some descriptive terminology to recognize the tendency to read bloggers with similar values/beliefs (cocoon) and to repeat the same ideas back and forth (echo chamber). What such environments make it difficult to do is surface information held by a few.

These are not new observations. In my duties teaching Introductory Psychology, I have had occasion to present research on conformity (Solomon Asch - for example see Bond & Smith (1996)) and to discuss problematic social phenomena such as "group think" (Janis, 1972). The concept of group think proposed that established groups may avoid topics that are uncomortable. The group functions to avoid conflict rather than in a rigorous fashion to carefully do assigned work. Ideas are often not challenged to see how well the ideas stand up to scrutiny and individuals may not offer criticism of an idea for fear the comments will be perceived as criticism of individuals associated with the idea. Groups do not often always use what group members individual know know in productive ways.

It is important to note that participatory environments function in different ways and characteristics of environments may influence what problems may be encountered. For example, consider differences between blogs and wikis. Each offers opportunities for interaction, but in different ways. With blogs, individuals exercise great independence in what they write and in what they read. There is the opportunity for interaction through comments, but if individuals read blogs of similar minded individuals the comments may be unlikely to offer very competing positions. In contrast, a wiki forces individuals into a common interactive environment. However, in such an environment, there is no guarantee that minority positions will even be expressed (as they might be in independent blogs).

Possible fixes

In a later comment, Surowiecki recognizes similar challenges and offers some possible solutions. He contends that:

Reading Surowiecki's suggestions reminded me of the generic suggestions for "brainstorming" I sometimes discuss as guidance for group decision-making. My simple description of brainstorming recommends the following steps:

The idea here is to at least reveal a diversity of ideas before strengths and weaknesses are debated.

So, there are challenges to a position assuming individuals within groups will benefit from what others know and that groups will move toward more carefully reasoned conclusions. There are also ideas about how known limitations might be addressed. We assume concrete ways of applying some of the proposed "fixes" will emerge as participatory environments receive more attention.

In this case, classroom and instructor facilitated projects may offer some advantages over self-initiated learning. While each of us may be motivated to feed our interests and biases, projects that require sharing of ideas within a designated group or perusal of material identified by an instructor may widen the range of ideas learners are expected to consider.

Need a concrete example/suggestion?

Let's take the example of blogs. The problem we describe here would likely arise when one limits blog reading to self-selected blogs. We may even contribute to this problem by explaining how to set up an aggregator to provide the recent posts from designated blogs. It becomes convenient and comfortable to read what the same people say day after day. If these writers have been selected because we share their perspective of values, we end up getting a steady diet that supports and perhaps radicalizes our beliefs rather than asks us to constantly evaluate why we take the positions we take.

Consider a method for diversifying the content you encounter. In addition to reading the thoughts of specific bloggers, create a feed of blog posts on topics of interest. For example, I and most educators have strong opinions on topics such as "No Child Left Behind", learner-centered instruction, multiple intelligences, brain based learning and many other controversial ideas and practices. I could read bloggers who share my perspective on these issues, but I could also create a blog feed that brings to my "reader" recent blog posts on any of these topics without attempting to control who authored the posts. Such an approach would expose me to a wider variety of perspectives.

See our description of blog topic search RSS feeds to see how this is done.

----

References

Bond, R., & Smith, P. (1996). Culture and conformity: A meta-analysis of studies using Asch’s (1952b, 1956) line judgment task. Psychological Bulletin, 119, 111-137.

Janis, I. L. (1972) Victims of Groupthink. Boston. Houghton Mifflin.

Sunstein, C. R. (2006). Infotopia: How many minds produce knowledge. New York: Oxford University Press.

Surowiecki, James (2004). The Wisdom of Crowds: Why the Many Are Smarter Than the Few and How Collective Wisdom Shapes Business, Economies, Societies and Nations New York: Doubleday.


| Home | Previous |