One of my recent posts on the unreliability of Wikipedia caught the eye of Edward Buckner, a medievalist, who shared with me a paper he has written about deficiencies in an Oxford University study of the reliability of Wikipedia.
You may be aware of a study in 2005 by the journal Nature that found Wikipedia to be, on the whole, about as accurate as the Encyclopedia Britannica. This study is what I take to be the frequently repeated claim by Wikipediasts that the two references are equally reliable, even though Britannica challenged the validity of the study.
A subsequent study bty Epic, an e-learning company, and Oxford University was published in 2012, and it is this study that Dr. Buckner addresses, particularly on the reliability of the Wikipedia articles on Thomas Aquinas and Anselm of Canterbury.
Dr. Buckner's paper, "Critique of the Epic/Oxford University pilot study into the comparative accuracy and quality of Wikipedia," catches some things that ther Epic/Oxford project did not: "It failed to pick up that the article ‘Anselm’ was one of the many articles in Wikipedia plagiarised from the 1911 edition of Encyclopedia Britannica in 2005. Thus, while it was meant to compare a ‘new media’ user-generated reference work with a ‘traditional’ reference work produced by selected paid expert advisors and editors, it was to a significant extent comparing an obsolete version of a traditional reference work with the modern version."
But that is not all: "More seriously, the study also failed to spot at least nine serious errors introduced by Wikipedia editors into the plagiarised section. One of the errors was the result of serious vandalism that has affected many articles in Wikipedia, still not cleared up."
And still more: "Failure to specify selection criteria leads to the risk of ‘cherry picking’ or ‘selection bias’ – choosing a sample that fulfils the desired outcome of the study, rather than one which is neutral and unbiased. The article on the late thirteenth century scholastic writer Duns Scotus was not selected, for example, despite being – on any measure of notability – one of the three most important philosopher-theologians of the middle ages (the other two being William Ockham and Aquinas, of course). The Wikipedia article on Scotus has been the target of vandalism, still uncorrected, and is patchy and uneven. The Britannica article, by contrast, is written by Allan Wolter, a Scotus scholar and author of many studies on Scotus. There is no comparison between Britannica and Wikipedia here. It is the same story for William of Ockham. The Britannica article is written by Paul Vignaux, author of Philosophy in the Middle Ages. The Wikipedia article is a mess, incorporating content from the Catholic Encyclopedia mixed up with random and often incoherent contributions from Wikipedian editors."
Thus you have, in just a couple of entries, most of the shortcomings that make Wikipedia unreliable: inadequate sourcing, erroneous information, amateurish and incompetent editing, and outright vandalism.
I don't propose to set up a clearinghouse here for errors and chicanery in Wikipedia; such an effort would turn out to be as compendious as Wikipedia itself. But if the inadequacies that Dr. Buckner exposes, along with others cited in previous posts, do not induce you to have serious misgivings about Wikipedia as a reference, then what would suffice?
Copyright © 2015, The Baltimore Sun