Politically controversial scientific topics like global warming and evolution are being ransacked on Wikipedia, leading to "egregious errors" that must be monitored. This was the finding of the first investigation into Wikipedia science edits of its kind, published in PLOS One, which recommends editors have their reputations quantified and highlighted on these 'controversial' pages. "As society turns to Wikipedia for answers, students, educators, and citizens should understand its limitations for researching scientific topics that are politically-charged," said coauthor on the PLOS One paper, president emeritus of the Cary Institute of Ecosystem Studies Gene E Likens. "On entries subject to edit-wars, like acid rain, evolution, and global change, one can obtain -- within seconds -- diametrically different information on the same topic."
Wikipedia edit wars are nothing new. There's even a tongue-in-cheek Lamest edit wars' page on the digital encyclopaedia. But Likens wanted to take a quantitative look at how this could impact, and potential damage, important science being shared on the seventh most visited site in the world. He also had something of a personal bugbear with the practice, having codiscovered acid rain in North America and established the link between fossil fuel combustion and the phenomena.
Although in his own world the concept of acid rain was not at all controversial, he found the Wikipedia entry on it was receiving "near-daily edits", with some of these resulting in "a distortion of consensus science". He kept tabs on this personally, but decided to look into how it impacted other areas of the sciences that can be politically charged: acid rain, evolution, and global warming.
Likens and coauthor Adam M Wilson looked at almost ten years of data on the edit history of these three topics from 2003, as well as those on the standard model in physics, heliocentrism, general relativity, and continental drift -- areas they knew did not carry the same political charge. None of the seven topics were scientifically controversial in the slightest. The pair looked at the daily edit rate of each page, including successive edits made by the same user; the mean edit size, calculated as the the total number of words altered on days with at least one edit; and the average number of page views per day for each.
They found, as was expected, that overall the three 'controversial' topics received far more edits than the noncontroversial. As an example, the standard model in physics had around ten words changed every few weeks, while the global warming page was changed two to three times a day with 100 words changed on average.
"The high rate of change observed in these pages makes it difficult for experts to monitor accuracy and contribute time-consuming corrections, to the possible detriment of scientific accuracy," the pair write in PLOS One. "As our society turns to Wikipedia as a primary source of scientific information, it is vital we read it critically and with the understanding that the content is dynamic and vulnerable to vandalism and other shenanigans."
As Wikipedia-use becomes the norm for many young researchers, these edit wars can present a genuine concern. But the remedy the pair suggests, is perhaps just as tricky for the open model the encyclopaedia is based upon. They suggest pages vulnerable to these kinds of edit wars be automatically flagged up -- a sensible and practical solution, but one that makes it impossible to trust a fact on the page. Second, they suggest user reputation be quantified and made visible "to help readers critically evaluate the content of a page". Again, this could be a practical remedy, if anonymised, but would likely place a greater burden on an institution run on donations.
All in all, the best advice the authors could give is for young scientists to ignore Wikipedia as a source completely (or at least study the sources list at the bottom of each article critically). But considering it is the seventh most visited site in the world, that is unlikely to happen, and also a great shame given the vast amount of good the service can provide in terms of offering free, abundant information.
This article was originally published by WIRED UK
