Sunday, October 03, 2004

Trusting Strangers

In the "real world," most of the sources we go to for products and information are known to us. They're either widely known institutions (e.g. NPR for news or Circuit City for TVs) or individuals with whom we have some relationship. However, in the online world, it would be very restrictive to only consider sources of either information or tangible goods that we already know. eBay, Craig's list, etc. often have products at much lower prices than even online retailers, and as our search assignment has demonstrated, in some cases very good information can be obtained from anonymous people on the web.

In our real-world interactions, we tend to use past performance to predict the future -- the so-called "shadow of history." If we've had a poor experience at a store or got bad advice from a friend, we'll be less likely to patronize that business or turn to that friend again. Online, however, we generally don't have any history with the people or sites with which we're interacting. So how do we go about deciding whom to trust in our interactions with online strangers?

Clearly, this problem is one of balancing risks and rewards. We could easily chose to limit our online shopping to and other brands we know, or ignore any information that doesn't come from major publisher. Doing so, we'd be secure in the knowledge that our online experience will be of at least a certain level of quality. But we'd miss the opportunities of getting better prices in peer-to-peer markets or of finding more current, thorough, or specific information from people not attached to large media company.

So how do we balance these risks and rewards? We all do it to some extent, often without thinking about it. Much of the information available to help make these decisions also come from strangers (e.g. retailer reviews like, eBay feedback, even Google PageRank uses the link structure of unknown websites to rate page quality). Are these sources sufficient indicators of quality? If so, how can we be sure that we can trust even them? What else indicates that an online entity is trustworthy? Or do we only superficially care and protect ourselves by limiting our exposure to risk? For example, I might be inclined to trust unknown websites' information when preparing a blog entry on baseball trivia, but less likely to do so in my thesis. If that's the case, how do we decide the threshold of acceptable risk?

I've been thinking about this topic a lot lately, both in conjunction with my research (quality filtering systems) and with the search assignment: I found a number of pages with thorough, well-researched information written by total strangers. Were I actually researching my search topic, I couldn't decide whether I would actually use the information or not. While all signs pointed to it being authentic, useful information, its lack of provenance was really setting off my paranoia about random information online.


Post a Comment

<< Home