Wikipedia has become a source of citations for scientific work, although scientists do not refer to it.

The study found that phrases from Wikipedia articles relating to actively developing scientific fields fall into scientific work.

Wikipedia solves table disputes and saves those who are trying to cheat at the evening of erudition. Quickly: in which country does the Nile originate? In which year did Gershwin write The Rhapsody in Blues Style ? In Wikipedia, you can find answers to all such questions - including scientific ones.

Wikipedia contains hundreds of thousands of scientific articles, and it provides a way to quickly refer to the molecular formula of Zoloft , the inventor of a 3D printer, and that the theory of tectonic plates is only about 100 years old. This site is a gold mine for science lovers, science bloggers and scientists themselves. But although scientists use wiki, they don’t rush to admit it. The site rarely gets into the list of citations, as a source, for example, of the history of studying the brain-intestine axis or the chemical formula of polyvinyl chloride .

But scientists rummage through the wiki just like all other people. A recent analysis found that Vicki’s relevance doesn’t lag behind the latest research — and that the words from her articles fall into academic work. These results not only point to the habit of using wikis rooted in an ivory tower . They also say that a free and accessible source of information plays a role in research progress, especially in poorer countries.

Teachers in high school, high school and college instruct students: Vicky is an unreliable source. It can be edited by anyone, and articles change from day to day - sometimes by one comma, and sometimes they are completely rewritten overnight. “Vicki’s reputation is unreliable,” says Thomas Shafi, a biochemist from La Trobe University in Melbourne.

But the same teachers — even college professors — while warning students of using Wiki, use it themselves. “Scientists are constantly using wiki because they are people too. That's what everybody does, ”says Doug Hanley, a macroeconomist at the University of Pittsburgh.

And, perhaps, the reputation of the site as unreliable is unjustified. Vicky is no less consistent than the British Encyclopedia, as a 2005 study published in Nature showed (the encyclopedia itself vehemently opposed such a conclusion). But no one dares to quote it as a source. “She is not respected like academic sources,” Shafi said.

Academic science may not respect Wiki, but Wiki itself definitely loves science. Of the approximately 5.5 million articles, from half a million to a million concerns scientific topics. And the constant addition of material by hundreds and thousands of editors guarantees the relevance of articles on a par with the most recent scientific works.

It is easy to track how the latest publications affect the wiki. They are in fact mentioned in the encyclopedia. But does this connection work in the other direction? Do scientific articles from Wikis seep into academic literature, even if they are not quoted? Hanley and colleague Neal Thompson, who studies innovation at MIT, decided to approach this issue from two fronts.

First, they identified 1.1 million of the most common scientific words in articles published in the scientific giant Elsevier. Hanley and Thompson then studied how often these words are added or removed from the wiki over time, and are quoted in the research literature. They focused on two areas, chemistry and econometrics , a new area developing statistical tests for economics.

There was a clear connection between the language of scientific work and the language of Wiki. “If a new exciting topic emerges, it leads to the creation of a new Wiki page,” Thompson notes. Then the language of this page is associated with later scientific work. Hanley and Thompson demonstrated that after the appearance of a new article in the wiki, later scientific papers contain more language similar to wiki than the language of scientific works in this area, published before the appearance of the article in wiki. There was a definite connection between the language of an article in a wiki and subsequent scientific work.

But was Vicki herself the source of this language? The study did not answer this question. It only observed how the frequency of occurrences of words in various texts increases. It could not prove that scientists read the wiki and used what they read in their work.

Therefore, researchers have created new articles from scratch in Wiki to find out whether their language will affect scientific literature. Hanley and Thompson commissioned graduate students in chemistry and econometrics to write new wiki articles on topics that were not yet on the site. Students wrote 43 articles on chemistry and 45 articles on econometrics. Then, half of these articles from each region were published in the Wiki in January 2015, and the other half were held as a control. The researchers gave the articles three months to penetrate the Internet. They then studied published scientific articles in the next six months in these two areas to search for specific words used in the wiki, and compared them with the language of articles that were not published.

At least in chemistry, new articles have proved popular. Published in the wiki and control articles were written on topics not previously covered on the site. These included records on the synthesis of hydrastin (the precursor of hydrastinin, a drug that stops bleeding). People are interested in articles enough to view them 4,400 times a month.

Words from articles leaked into scientific literature. Six months after the publication of an article from Wiki, about one word out of 300 influenced in new scientific works in chemistry. And research papers on the topics described in Wiki, over time, began to look more like articles in Wiki. For example, if chemists wrote about the synthesis of hydrastin - on the topic of one of the new articles in Wiki - in the published scientific papers the phrase “Passarini reaction”, the term used in the article in Wiki, was more and more often encountered. But if the article was not published in Wiki, then the scientific works did not become similar to such an article (and this could happen simply because this topic was gaining popularity). Hanley and Thompson published a preprint of their work in the journal Social Science Research Network on September 26.

Unfortunately, the articles on econometrics in Wiki did not trigger the reaction. “We wanted to choose something on the verge of science,” says Thompson. But the line turned out to be too thin. New articles on econometrics in Wiki were viewed thirty times less than articles on chemistry. Thompson and Hanley could not gather enough data from the articles to make reasonable conclusions. Well, we wish econometrics good luck next time.

Linking articles on wikis and scientific literature differed in different regions. When Hanley and Thompson sorted out published scientific articles on the GDP of the countries from which they originated, they found that the wiki articles had a greater influence on the language of scientific work of scientists from countries with weaker economies. “If you think about it, then in a relatively rich country, a person should have access to a large number of journals and scientific literature,” says Hanley. And organizations in poor countries can not afford expensive subscriptions to scientific journals, so scientists from these countries have to rely more on publicly available sources, such as Wikis.

This Vicki study is distinguished by “a well-designed job scheme and robust analysis,” says Heather Ford, who studies digital politics at Leeds University in England. “As far as I know, this is the first work indicating a strong link between articles from Wikis and the development of science.” But, as she notes, all this concerns only chemistry. In different areas and influence may differ.

“She addresses a question that has long been of interest to people, which is difficult to prove,” says Shafi. There is a clear connection, but, according to him, keeping track of the language is not the same thing as keeping track of how new ideas and concepts migrate from the wiki to the ivory tower. “To say that more research is needed on this topic will certainly be a cliche, but in fact it is the way it is”.

Hanley and Thompson would agree first. “I think this is the first step,” says Henley. “It is shown that Wiki is not just a passive resource, it affects the cutting edge of knowledge.”

This is a great reason for scientists to get involved in the work and edit articles, Thompson notes. "This is a major scientific resource, and it seems to me that we must recognize this," says Thompson. “It is useful to make sure that the science in the wiki is well and fully described.” Good science wiki articles can not only solve table disputes. They can help advance science. After all, in the end, scientists are watching her, even if not admitting it.