Reliability Requires Redundancy

from the peer-review dept

Slate has an interesting column by Jacob Leibenluft that compares the relative reliability of Yahoo! Answers and Wikipedia. He notes that unlike Google Answers, which was shuttered last year, Yahoo's answers site has become quite popular. Google paid its contributors to answer questions, but Yahoo! only offers contributors points that entitle them to ask questions of their own on the site. Unfortunately, even the service's own users acknowledge that Yahoo! Answers isn't very reliable. Often a question will attract a dozen or more answers. Some of them will probably be right, but others will be wrong, and it might be hard to tell which is which. Leibenluft contrasts that to Wikipedia which, while far from perfect, tends to have accurate information the vast majority of the time. He zeroes in on three important characteristics that give Wikipedia the edge. First, whereas Yahoo! Answers lets each user to give his or her own answer, the Wikipedia process is geared toward producing a single article that represents the consensus of all contributors. Second, Wikipedia has a strong norm of citing outside sources when contributors disagree. Usually, when there's a factual disagreement, someone will go out and find a citation in a reliable source to demonstrate the correct answer. Finally, Yahoo! Answers closes a question after about a week, whereas a Wikipedia article is open for editing indefinitely. This is important because Wikipedia articles tend to get more accurate over time, as more and more readers visit them and fix mistakes. It would be interesting for Yahoo! to experiment with a wiki-based format for Yahoo! Answers, where users collaborate on a single collective answer to the question rather than giving a bunch of individual answers. The major difficulty would be that the site's point-based reward system would be difficult to apply, since several users would have contributed to the final answer.

Techdirt's own Insight Community is similar in some ways to the Yahoo! and Google Answer programs. The failure of Google Answers might be a reason for pessimism, but I think there are a few key differences that make TIC more likely to succeed. First, the community is sharply focused on a fast-changing industry where expertise is especially valuable. Second, TIC is focused on providing insight and analysis, not just plain facts. With factual questions, a customer will typically be seeking a single correct answer. But with strategic business questions, there usually isn't one right answer; companies are often interested in hearing about several different approaches, and there can be a lot of value in seeing the arguments that experts marshal for various options. Finally, Techdirt is much more selective about the experts it brings onboard, using experts' blogs and other writings as a way of identifying those who know what they're talking about and can communicate it clearly. That gives the TIC a great signal-to-noise ratio.

Filed Under: facts, google answers, insight, redundancy, wikipedia, yahoo answers
Companies: google, techdirt, yahoo