According to Wikipedia, a megajournal is, “a peer-reviewed academic open access journal designed to be much larger than a traditional journal by exerting low selectivity among accepted articles.” The term itself has an unclear heritage, but was promoted memorably by Peter Binfield when he was at PLOS. Some say Binfield coined the term. Over the years, the “rise” of the megajournal has been a consistent source of excitement or concern, depending on your perspective, and the apparent success of the megajournal model has attracted a number of imitators and competitors.
Toward the end of March, I decide to tour a few of the megajournals that have cropped up over the last decade in the midst of many older homes.
This tour was done more out of architectural interest than anything. Just as McMansions (extra large houses with extra rooms and many large rooms) emerged as a fad in real estate, and only later were understood to represent a long-term trend and preference as “open” design plans and larger houses became popular – with kitchens, family rooms, and dining rooms barely distinguishable – megajournals are our new(ish) kids on the block. Do they represent a long-term trend? What makes them different? And which parts might persist beyond the fad stage?
Caveats for this tour abound – I did not spend more than 30 minutes on any one site, I visited each site only once (so that day’s experience is only a sample of the overall experience), and I didn’t tour all of the megajournals. I limited my tour to four that come to mind when I think of the category – PLOS ONE, Scientific Reports, Nature Communications, and Heliyon. I drove by a few others on the trip, so they get some mention.
Here’s what I found.
Big variations in the volume of articles. PLOS ONE still takes the crown for volume of articles, publishing 100+ papers per day. Nature Communications and Scientific Reports are also publishing a lot of papers, but only about 10-20 per day. It appears Heliyon is struggling to meet the megajournal pace, sometimes publishing nothing for days at a time. But it’s new, and entered the game late. Maybe too late? One big question is about what a huge volume of publication means for a journal and for a community. A recent editorial in the European Spine Journal despaired about the rapid influx of papers of limited utility or low quality. In nearly the same breath, the author mentioned a seminar in Asia to help authors get their papers accepted in the European Spine Journal. It seems everyone is seeking papers, even if they’re feeling conflicted about the situation. But the effect of more papers seems to be neutral at best. In the megajournals, a scan of their article-level metrics indicated that most articles aren’t being read much soon after publication, which suggests low initial awareness of their existence to relevant audiences, or very small relevant audiences, or some combination of the two. Email alerting services across the megajournals were uneven, with the Nature properties doing a better job, and PLOS ONE offering options I had a hard time finding and hesitated to use, they were so vague about what I’d receive. Given the volume of articles, even the useful email alerts I’ve received are pretty overwhelming.
Home pages are minimally curated, have different purposes. In most of the cases, home page design seemed focused more on presenting a landing page experience geared toward author marketing than a content experience geared toward readers. Large “Submit” buttons and solicitation zones dominated many of the home pages of the megajournals, with Heliyon and Scientific Communications being especially noteworthy. In Heliyon’s case, a “Submit” widget followed me around the site – which is interesting, because of all the sites, Heliyon seemed the least like a megajournal when it came to pace of publication (not as many articles being published yet). Across the set, home page content was usually presented as a reverse-chronology list. For PLOS ONE, on the day I visited (March 22, 2016), the fact that the home page article set for Recent Articles was driven by a search algorithm meant that every article on the home page was a Correction notice. Six corrections had been published at the end of the previous day, apparently. This de-emphasis of design and effort around the home page makes absolute sense, and is something I wish traditional journals would get their heads around – home page traffic as a percentage of overall traffic is falling, content discovery via the home page is a diminishing traffic driver, and editorial/design time spent on perfecting the home page can seem an exercise in vanity. Google and CrossRef and social media have really driven us into the article economy when it comes to discoverability, and home pages don’t matter as much. One other journal that received a quick look for the home page experience was SAGE Open, which had an especially stark design and implementation, with a utilitarian layout and searing white background.
Article designs aren’t as developed as you might imagine. You would think article design would be a strength of the megajournals. I didn’t see any sign of obvious techniques to improve readability or discoverability. Overall, the designs are very similar – single-column with section links for jumping around, a prominent PDF link, and some sharing tools. Then again, why would you expend resources on readability and discoverability when the business incentives are around more submissions? Even sharing tools were generally poorly implemented – a common design condition on many journal sites – and the areas in the rails (right or left) were the usual jumble of what always seems like compromises from meetings. Going to the PDF, most megajournals provided a simplified document with a single-column layout and embedded figures. Nature Communications provides a two-column PDF, which fit the more traditional feel the journal has overall. The single-column PDF is nice enough, but hearkens to the manuscript. This approach is becoming more common as production costs for 2-3 column PDFs have driven more publishers to the single-column solution, which is cheaper to automate. In the case of Heliyon, the PDF felt very much like a gussied-up author manuscript. Design at Heliyon overall (site, usability, PDF) was underwhelming. Despite a lot of talk about article-level metrics at the megajournals, these were also a bit spotty – not all that clearly available, updated somewhat idiosyncratically, and not as well-presented as those I’ve seen on some traditional journals.
Publication and production times aren’t all that quick, seem to trend toward the mean. It’s common for submission, acceptance, and publication dates to be listed. Overall, it seems that it takes each journal two weeks from acceptance to publication, but times between submission and acceptance weren’t all that quick, often months in length. Checking on this beyond sampling and a general impression would require a robust study, which is entirely possible. This is not that. But my brief review left me with the impression that these journals are taking months to accept submissions, and two weeks to produce articles from accepted manuscripts. Given what is an increasingly common and commoditized production and review environment (these journals likely outsource various functions to some common service providers), the notion of an emerging de facto standard for review and production times may not be that surprising.
More regional science and findings. One striking thing across the titles was the number of trials that have regional flavors. This was most apparent in PLOS ONE, but it existed elsewhere. Swaziland. Eastern Morocco. The Maldives. Egypt. Southeast Asia. South America. I can’t comment on the merits or value of these studies – small ecosystems and local populations can hold great interest for science and medicine – but the ratio of these seemed higher than I’ve seen elsewhere, and suggested niche scientific and regional audiences.
Subject and topical execution varies. One challenge for megajournals is categorizing their content so that it’s usable for researchers in particular fields. It’s not a challenge they’ve conquered in any innovative way, really. Their approaches seemed pretty stock, and in one case, downright puzzling. Some divide their content up better than others. Scientific Reports seems to have done about the best job of the set, allowing initial searches to be categorized easily and using tagging well across the site. Heliyon does a poor job – their content tags on articles aren’t even hyperlinked, which was the head-slapper in the group. PLOS ONE is mediocre in this regard, with decent tagging that seems a little overcooked, to a search engine that I recalled as being better a year or two ago than it is now (it seems they tried to simplify it, but instead made it into a simpleton). PLOS ONE has a feature that I doubt anybody uses – tags that lets users indicate whether the tag was useful or not.
Technical execution varies. The megajournal sites I visited varied quite a bit in obvious technical execution. Of the set, Heliyon was the laggard. The site felt like it wasn’t fully baked, and usability problems were pretty obvious (example: if you have the search window active, you can’t access the links to download the PDF or use other article-level services; ). At PLOS ONE, a persistent message of “Loading metrics information . . . “ appeared on each article as listed in summary form, as if the system were struggling to access a related database. Scientific Reports was the best of the sites, with snappy performance, clean design, strong usability, and no obviously broken features. Nature Communications wasn’t far behind.
The submission experience. I couldn’t go far into the submission experience without creating bogus credentials and uploading a bogus paper (not that this would represent entirely new territory). In some cases, calls to action for submissions were very aggressive (Scientific Reports and Heliyon). Across the megajournals, there were some noteworthy differences. PLOS ONE has a “Publish” item in its headers, and this leads to a massive dropdown menu that, once you find the “Submit” entry, leads to a landing page, which forces another click before landing on a branded Editorial Manager frontend. For Heliyon, you’re redirected to EVISE (funny name), the Elsevier submission system. Heliyon also has the widget that chases you all over the site asking for submissions. Both Nature Communications and Scientific Reports resolve to the same submission form (except for different branding), but taking the Scientific Reports path forces you through a landing page first, with the link to the submission system a little hard to find, and a more prominent “Publish >>” button that oddly makes the landing page reload instead of taking you to the submission system. (It’s also odd to call a “submission” button “publish,” but I’ll let that little bit of semantics speak for itself. PLOS ONE does the same thing.)
Impact factors. In addition to touring the sites, I looked at the impact factors for the titles that have them (PLOS ONE, Scientific Reports, and Nature Communciations). Heliyon is too new. Nature Communications, which struck me as the most like a traditional journal in its design and “feel,” has the highest impact factor (2014 = 11.470). This may be partially due to the fact that it carries the Nature brand, which has been shown to be a powerful attractions for authors and citations, as well as pointing to an impressive publishing infrastructure. Scientific Reports is an unbranded Nature journal, with an impact factor (2014) of 5.578. Both Nature megajournals reside within the www.nature.com domain, a discoverability benefit, I’m sure. As mentioned above, the Nature titles also had better email alerting facility. Since citation begins with awareness, this may be an element of their impact factor success. PLOS ONE trails the group with a 2014 impact factor of 3.234. But what truly jumps out is the vast difference in published articles between these three. For the prior two years, PLOS ONE published 54,945 citable objects (scholarly articles), while Nature Communications published 2,297, and Scientific Reports published 3,278. Together, the two Nature “megajournals” accounted for just slightly more than 10% of what PLOS ONE alone published. What is the threshold for “mega” these days? More on this later.
Odds and ends. Abstracts in PLOS ONE varied, from structured (mostly biomedical papers) to unstructured. And the structured ones varied in structure. It seemed almost as if each domain culture imposed its abstract habits on the journal. This may speak to the lack of curation and form imposed by some megajournals. SEO for all the journals seemed decent, with Nature’s titles even reconciling one of their so-bland-it’s-confusing titles properly (e.g., search for “Nature Reports,” and you get Nature Communications; but if you search for “Scientific Communications,” you don’t get Scientific Reports). Not a lot of care was taken with the tone of article titles. For instance, in a few cases, I was drawn into a negative trial because the title was written as if it could be a positive trial. While still a fan of negative trials in theory, in practice I think they have a higher bar to generate interesting results, and these didn’t seem to hit those heights. Certainly, an editor could have demanded a title that wasn’t slightly misleading. It made me wonder if some of the suggestive language authors can sneak into papers to goose their results’ apparent importance was also addressed during editing.
So, what does it all mean? After touring these sites, and thinking about why they emerged, how they fit into the overall scholarly information economy, and how that economy itself has changed in the last few years, I am left with three overriding thoughts.
What is a megajournal? We may want to rethink the use of the term “megajournal.” While these journals may have been designed to be much larger than traditional journals, some clearly are not meeting the standard. Except for PLOS ONE, they aren’t that big. For example, Nature Communications and Scientific Reports published fewer articles in 2012/13 combined (5,575) than the Proceedings of the National Academy of Sciences (7,704) or the Journal of Biological Chemistry (7,321). Does a journal have to be OA for it to be “mega”? It’s interesting to contemplate that a journal format specifically designed to be boundless has apparently only once truly achieved what the “mega” in its name suggests. We might also have to reconsider whether there is a “rise” of megajournals, or whether they are reaching the end of a growth phase in both quantity at the journal level and number of titles with megajournal characteristics overall.
Article commoditization. Whether or not there is truly more than one megajournal, the concept represents part of a strong trend toward the commoditization of articles by publishers in response to the atomization of the journal by search engines and social media. Megajournals are not the only hallmark of this trend in publishing. You see a different approach to commoditizing articles with the portfolio approach of the bigger brands, with cascading journals and so forth. Whether a brand captures thousands of papers in a system of journals (is this a “megaportfolio”?), or a single journal captures thousands of papers in one title, seems to matter less and less. The competition starts with papers, and more publishers are entering the fray, making papers more of a commodity.
Whither the reader? A lack of focus on the reading experience is a glaring Achilles’ heel of the commoditization of articles. In many fields and at many career stages, readers still respect, seek, and want a curated experience. They understand and appreciate how editors string related content together, surround important new research with editorial perspectives from respected peers, set the pace and priority in a field, and create hierarchies of content to suggest consonance or dissonance among ideas.
I’ve been speaking with scientists and physicians quite often lately. Listening to them, you realize that their relationships with journals, as readers, can be dramatically different from how we’re treating journals these days. These relationships are often more emotional, long-lasting, and essential to their identification as professionals. While they use Google and PubMed and Scholar, they often put journal brands into their searches. And many don’t search online for content. They take what they get, adopting that “if it’s important, it will find me” attitude. But are we doing enough to set content on that course?
The move into treating articles as a commodity seems mostly due to economic necessity – i.e., since price increases are off the table by and large, and the volume of papers continues to increase, the economic option that remains is to increase publication volume. So, we’re making and selling more articles. Meanwhile, readers are largely being left out of the equation. This is the hidden cost of a neglected information economy.
On the bright side, these journals are serving a community need – there are too many scientists who need to publish, and providing outlets helps solve their immediate problem. But it still feels rather cynical, as the long-term issue with what these represent – high-volume publication services for authors who value speed and convenience – is that we are losing sight of the readers, a critical connection for authors, editors, and publishers. And a critical connection for scientific advancement.
My overall view of the megajournals I looked at boils down to this:
- Not nearly as innovative with technology or UI/UX as I would have hoped
- Weak implementations of social sharing technologies (they are not alone)
- Strikingly similar to one another in some ways – home page approach, PDF design, use of author appeals, publication and production times, etc.
- Difficult to use as discovery platforms, but likely effective enough in search engines
- Not something I’d seek to read, as the relevance of the content is really uneven and unpredictable – these are not created for anyone, but represent decent services to time-starved authors