By Brian Owens
A few years ago, Jorge Castillo-Quan, a post-doc studying the biology of ageing at University College London, UK, wrote a report about how insulin and cortisol interact to affect brain function. It didn't seem to make much of an impression on the scientific community, and was not highly cited, so Castillo-Quan moved on to other topics. But recently Castillo-Quan has seen a sudden resurgence in interest in his work on cortisol, in the form of people reading and downloading the paper from his profile on the academic social networking site Academia.edu. “The cortisol work has been seen by probably 50–100 people so far this year”, he says. “And they're not looking at my newer work.”
The attention has come from various researchers around the world, he says, indicating that the topic might be becoming interesting again to the field. Now because of that resurgence in interest Castilllo-Quan is considering whether he should go back and perhaps pick up where he left off. “I'm looking at what would be interesting to revisit”, he says.
This kind of story is remarkably common among users of Academia.edu, says Richard Price, the site's founder, who received a PhD in philosophy from the University of Oxford, UK, before starting the site. Many users have told him that they use Academia.edu's rapid feedback metrics on the number of views and downloads of their papers to decide where to focus their efforts. “It's a shorter feedback loop than with citations”, he says.
Facebook for scientists
Academia.edu is just one of several sites to crop up in the emerging landscape of academic social media over the past several years. A similar service called ResearchGate was started in 2008—the same year as Academia.edu—by Ijad Madisch, a former physician and virologist in Berlin. And Elsevier, the company that publishes The Lancet, bought Mendeley in 2013, a site that originally focused on managing and storing documents but also encourages social networking. Alongside these are the remains of several earlier, failed attempts to create a “Facebook for scientists”, including Nature Network, by the Nature Publishing Group.
Both Price and Madisch started their social networks because they were frustrated with the difficulty of sharing the results of their research through the traditional academic publishing channels. Other researchers don't always have access to all the same journals, and it can take a lot of time and effort to build and update a personal homepage. “I didn't want to just post my work and hope people would Google it”, says Price. “So it was natural to think of a network.”
Madisch was particularly concerned that scientists were not sharing all their results in the most efficient way. “Tons of negative results are not shared, so people keep making the same mistakes”, he says. “We would make more progress if we published all of our results.” Social networks like ResearchGate provide a way to do just that.
These academic-focused social networks operate much like their more traditional counterparts such as Facebook or LinkedIn. Members fill in a template to create a homepage with their CV, research interests, and a photo and link to or follow other scientists or topics in which they are interested. But instead of posting baby pictures or cataloguing the minutiae of your day, they upload copies of their latest research publications, and discuss the technical aspects of their work.
The sites have done a good job of signing up users. Mendeley has around 3 million members, ResearchGate has 5 million, and Academia.edu has around 14 million—although that higher number is at least in part due to the fact that Academia.edu allows anyone to join, whereas ResearchGate restricts membership to those affiliated with a recognised research institution. Madisch says that restriction is necessary to keep the quality of discussion high (although the data and papers on the site are available to everyone), whereas Price says his more open site allows for interesting and unexpected collaborations—such as a film director seeking technical experts for his project.
The members of these sites tend to be younger than the overall scientific population—most are PhD students or post-docs—probably because they are more comfortable and familiar with social media, and are used to using technology to find and make connections with people who have similar interests. But Madisch says the number of more senior scientists on ResearchGate is on the rise, and proudly says that one of his former supervisors, who had originally discouraged him from starting the site, recently set up a profile.
It's not clear, however, how many of these members are really active on the sites. A survey by Nature, published in October, found that although academic social media sites enjoyed high visibility among researchers—more respondents said they were aware of or visited ResearchGate regularly than Facebook or Twitter—most said they mainly used ResearchGate or Academia.edu simply to maintain a profile in case someone wanted to contact them. Posting and finding papers, discovering related peers, and tracking metrics were the next most popular uses, but were far less common.
But activity on both sites has been growing fast. Madisch says that about a third of ResearchGate users log in to the site at least once a month, and those who signed up within the past year are much more active than older users. And both Madisch and Price have no trouble finding examples of members who have used the sites to find new collaborators, solve experimental problems, or show the international reach of their research to tenure committees.
The number of papers and datasets uploaded is also mounting at an exponential rate. “In the first 4 years, people uploaded about 2 million publications to ResearchGate”, says Madisch. “Now we get 2 million every 4 weeks.” And researchers are adding raw datasets and negative results at a rate of 1300 per day, he adds.
The number of research papers on Academia.edu is also expanding rapidly, with 15 000 new papers uploaded each day, says Price. As a result, he says, Academia.edu's Facebook-like news feed is a “gushing firehose of papers in your field”. That firehose has proved popular with researchers, especially those in developing countries who often do not have access to the latest publications through their institutions. But publishers are less impressed. Several have sent takedown notices to the sites in the past year, asking them to remove their copyright material. Madisch is not concerned, though. “Scientists have contracts with the publications, and they have to follow the terms. We have to take them down when asked”, he says. But in most cases the copyright infringement can be avoided as long as the author uploads the final accepted manuscript, rather than the actual journal article. “When we started, the idea of sharing papers was not entirely mainstream”, says Price. “But now that the concept of open science and open access has more support, it's irreversible. Soon we'll get to the point where all papers will be open access.”
The future of research
Both Price and Madisch have high hopes that their online communities can change the way researchers work and communicate their results. Price says that the biggest effects will be felt in how peer review is done. “I'd like to see a world where peer review is much faster and lower cost, so we don't need paywalls”, he says. He also wants the system to be better at filtering out non-reproducible results.
How, exactly, sites like Academia.edu can make that happen is, he admits, “the $10 billion question”. But it's fertile ground for exploration. “We're about to see a 5-year period of intense innovation around peer review”, Price predicts, “analogous to where we were 6 years ago with open access”.
Some researchers are already taking that next step. Earlier this year Frederik Feys, a clinical sexologist doing a PhD on the placebo effect in sexual health at the Free University of Brussels, submitted an article examining the use of the drug dapoxetine to treat premature ejaculation to the open-access journal PLOS One. Weeks later, the journal came back saying it could not find enough reviewers. “I just thought, ‘what am I doing’, the specific aim of this work was to try and get results out quickly”, says Feys. So he posted the manuscript on ResearchGate, and invited reviews. One reviewer replied quickly, before the staff at ResearchGate noticed his experiment and helped find more. Soon he had three reviewers, which he says, helped to make the manuscript much better. On Oct 7, Feys became the first scientist to publish a paper solely on ResearchGate. The entire process took around two and a half months. All of the reviewers comments and each draft of the manuscript are available for anyone to download and comment on. So far, the final paper has been downloaded more than 160 times. Feys says using ResearchGate as a publishing platform had several advantages, including being completely free, and taking far less time than a traditional publication. He experienced no problems with having the peer review be done openly.
The disadvantage, though, is that the credibility of his work is lower when it is published in a non-traditional way. “I'm a PhD student, my supervisor expects me to publish in traditional journals”, he says. So after an update, he plans on submitting it again to a regular journal. “To be in the scientific mainstream, you must be in the usual journals”, he says.
Madisch says Feys' use of ResearchGate as a publisher is the “next big step” in what the site can do for scientists. And, he says, it shows how scientists can use the internet to innovate and improve how their work is shared. “Journals are an important part of scientific life, but they haven't adjusted to the online world”, he says. “The world wide web was developed by scientists for scientists, and it has expanded to improve how almost everything is done, but science is still done the same way.”
Article source: The Lancet. Published on November 22, 2014.