Reliable knowledge

I've often wondered, given the amount of free information available through Wikipedia and other internet sources, whether it's worthwhile subscribing to an on-line encyclopedia, such as the Britannica. Would it be more balanced, better researched? (A second but equally important issue is whether it is better presented, and, hence, a better resource for children's homework than the rather dry presentation of Wikipedia?)

A report published by Nature at the end of last year suggested a similar level of inaccuracies in both Wikipedia and the on-line Britannica. This has yielded, understandably, a vigorous riposte by the publishers of Britannica. Having been brought up with Arthur Mee's Children's Encyclopedia as a standard family reference I've always been a little sceptical about the authority of encyclopedias. During the space race of the 1960s we took great (and unfair) delight in Mee's assertions that space travel was an impossibility. The inaccuracy didn't stop us enjoying the encyclopedia (although the search from index in volume 10 to multiple volumes of the encyclopedia itself did, but that's another story of a bygone paper age).

Coincidentally, economist and columnist Tim Harford reflects in this weekend's Financial times on the reliability of book reviews on Amazon, and on their impact on buyers' behaviour. According to research at Yale School of Management bad reviews have more impact on buyer behaviour than good ones, possibly because review readers tend to perceive good reviews as hype from the author's friends. I suppose the informed reader will apply a similar filtering process to the Wikipedia biographies of American politicians.