Thursday, October 28, 2004

Don't trust me: I make use of people

Slightly embarrassed to be writing a posting on trust… something I thought I knew a lot about, but was apparently wrong. But an assignment is an assignment, so here goes:

Title: Trust in online/face2face interaction, & Reputation Management Systems

Think about the popularity of fan fic, blogs, Usenet, forums, etc. for a minute. Why do you subscribe to certain writers, bloggers, Usenet newsgroups, forums? Like it or not, there is a certain level of ‘trust' you have in that medium that keeps you going back, even if you don't like that particular person's agenda (e.g. Bush or Michael More haters). So, ‘trust' in an online environment is something worthy of discussion.

Dealing with trust in an online (or in a face2face) environment boils down to past interactions with the other party. More accurately, it is the history of past interactions that gives you a better representation of that party's abilities and trust worthiness (Resnick, 2000, p2). It does not matter if we have never dealt with that person, so long as a clear history of transaction by others who had dealt with that person is made available to us.

Reputation management systems in an online environment give us this opportunity of a ‘third eye' to evaluate people based on the reviews of others who interacted with them. Resnick (2000) believes that for reputation systems to work there have to be 3 properties in place:

  1. longevity of entities

  2. current feedback is captured and distributed transparently

  3. Past feedback posted to guide buyer decisions

Longevity of entities ensures that future interaction will occur. This creates expectation and adds to the belief that the reputation system works. A transparent system of past and present feedback will help others who have never dealt with him/her before to paint a picture of that person, and better gauge their credibility.

In theory, Economists will point out that in eBay's system where feedback is overwhelmingly positive, that a potential ‘lemons market' could develop, whereby some sellers can take advantage of the system of predominantly trust and good relationships to sell some poor quality products.

In practice, this is not the case. So what prevents a lemons market from developing?

Kollock (1999) likens informal online markets to the financial exchanges that happen in wall street stock markets. Basically both environments offer conditions that promote trust among buyers and sellers. The environment has to have 1) matching service, to pair off buyers and sellers; 2) method of information exchange, where past and present transactions are recorded, and distributed; 3) a clearing service, (e.g. escrow) where by transactions can be completed safely.

Stop and think

Let's suddenly bring this whole discussion to a halt, and now put it in exclusively in a Face2Face environment. What changes? … essentially nothing. Resnick (2001) points out 8 factors that affects trust in traditional exchange of goods. Caixia's blog actually retyped out these 8 factors (thanks!), and I agree with her that Reputations are more important in facilitating cooperation online than offline.' Dave wade is another supporter of this belief. However, I still fail to see how these 8 factors are related in an online environment.

What my classmates had to say…

I liked Marion's post on trust where he quoted from his own experience with his class where there was no cooperation among students and ultimately did not obtain the highest payoffs possible. It got me thinking of the following scenario:

Think about a situation in a classroom (face2face or online, it does not matter) where students are asked to peer-review each other. For those who know Prisoner's dilemma or Game theory (call them bonnie & Clyde), they will rate everyone in the class low, on the assumption that everyone else will rate each other fairly. This will result in bonnie & Clyde scoring significantly higher scores than everyone else . When high assessment stakes are on the line, what's trust got to do with anything?

Deonne's post approached it from a different angle, and asked a series of good questions to evaluate reputations in both online and offline environments, namely:

  • Are good reputations necessary? Rewarded?

  • How are reputations established?

  • How does reciprocity fit into the cycle?

  • What are the benefits of having a good reputation?

  • Are people honest when talking about others?

  • Should censorship be allowed?

I like this particular posting, probably post of the week among those I read. I agree with Deonne that in setting up online reputations like in eBay, ‘Newcomers are typically distrusted until they “prove” themselves.' It goes along with the critical mass some sellers/buyers attain after sometime where feedback becomes a norm for them, and they don't even bother to respond to negative feedback anymore.

The biggest gripe I have with offline reputations is the tricky business of eliciting feedback. Because in face2face situations we do not have the protection of the anonymity of the web. Deonne's example of students filling out class evaluation forms in the U of U is an excellent example of avoiding possible repercussions of honest feedback. And we all know how valuable those information are!

I did not really think of censorship in depth until I read her post. To me, this brings a new equation to the game. Censorship can determine what kinds of comments get posted, and who gets to post. Same thing applies to blog post & comments too.

For example, if we housed a blog system in a school, then technically all those post & comments belong to the school. Any repercussions that result from controversial post/comments will be directed to the educational institution. Who then determines what types of religious, stereotyping, political, other sensitive issues are to be considered taboo subjects? What about countries without first amendment right protection or freedom of speech? How is censorship different in those context? I do not the answers either, but like the observation Deonne raised.

Edisoabi offered some possible tools ( blogdex and Daypop ) to start researching on trust in blogging communities. Lastly, Curtis' blog beautifully uses the ‘santa-claus-principle' analogy to illustrate why eBay's system works even when people don't know how it works.

Considerations of an internet-based reputation system

What are the incentives & disincentives of providing good & bad feedback in an online environment?

Talk about blogs. You are a blogger. You want traffic. So, you read others blogs and post positive comments hoping that it will make them like you enough to read your own blog. Would you curse and swear at their blog, and say they are posting garbage even if 90% (my own biased estimate) of blogs out there are? Will you be not afraid of the repercussions out there? It's the same thing with how sellers, buyers work on eBay.

Resnick (2001) coined ‘high-courtesy-equilibrium' to explain the lack of free riders and rarity of neutral or negative feedback on eBay. eBay has created an environment that promotes people to give positive feedback, and hence spurn more transactions. The commenting system on eBay is flawed due to this biasesness, but it works.

Is one good feedback = one bad feedback? I think the challenge now is to balance the weights of the rare (but sought-after genuine negative feedback) as opposed to the avalanche of good feedback that surrounds us. Same thing goes to the post and comments in blogs/forums, etc. Closest thing I see to balancing this now is probably Slashdot . But that is for another discussion.

In a nutshell : it's not how the system works, but that its participants believe it works – even if they don't know why (Resnick, P., Zeckhauser, R., 2001). In an extreme sense, ‘Blind faith'.




Kollock, P. (1999). The Production of Trust in Online Mark ets. Retrieved on October 24 th , 2004 from source:

Resnick, P. (Dec, 2000). Reputation Systems. Retrieved on October 24 th , 2004 from source:

Resnick, P., Zeckhauser, R. (2001). Trust Among Strangers in Internet Transactions: Empirical Analysis of eBay's Reputation System. Retrieved on October 24 th , 2004 from source:


At October 28, 2004 at 10:12 AM, Blogger David said...

Absolutely smashing post. 20 points. The linking and referencing shows you really starting to mature in the genre. It's really cool to see.

At November 2, 2004 at 12:23 PM, Blogger JoseGomez said...

Great post BH.

I still not convinced that informed consumers are satisfied with blind belief. I think as consumer become more experienced and aware of the dangers of the internet (esp. those involving monetary transactions), we do become cynical and do not just conduct themselves on blind faith but with a critical mind. But I could be wrong.



Post a Comment

<< Home