Guest post by Mark Ware, a consultant who assisted with the analysis of this years’ Author Insights Survey. The data from this survey has been made open access for anyone who would like to use it.
It used to be said that journal publishers knew little of researchers and the detail of their roles as authors and readers, despite representing the lifeblood of their businesses.
This has changed dramatically in recent years, with a series of large-scale surveys from publishers and academic researchers, combined with other techniques including online log analysis, focus groups, usability labs observation and ethnographic studies.
The latest Author Insights survey from Nature Publishing Group continues this trend and provides a host of fascinating insights into researcher attitudes. The survey repeats some questions from their 2014 and 2013 surveys, including open access topics, allowing changes to be tracked, but also covers the new topics of journal reputation and how researchers value publisher services.
Over 22,000 researchers completed the survey, with the analysis based on the 21,377 respondents that had published at least one paper in the last 3 years. These active authors included researchers from STM and HSS (humanities and social science) fields, though numerically dominated by STM respondents (86%).
The extent to which open access has become normal is shown by the 59% of researchers that had published at least 1 OA paper in the last 3 years (47% with an APC; 27% without). This proportion only grew a little compared to 2014, though, with what growth there was concentrated in HSS.
A small minority (7%) published exclusively OA; these fans tended to be younger and more likely to be based in Asia (13%).
On the other hand this still leaves 40% not publishing OA, rising to 43% in the physical sciences and 58% in HSS. This wasn’t because authors were unaware of OA options: only 6% cited this as a reason. The highest-ranked obstacle to publishing OA was concern about the perceived quality of OA journals, showing this issue remains potent, particular for older researchers and those in N America. After quality concerns, cost issues dominated, with around a third saying they were unable or unwilling to pay APCs.
Funder policies are acknowledged to be a major influence on the growth of open access, but researchers in this survey were poorly informed about their funders’ requirements. Some 35% said their funder made no requirements, but analysis of these respondents’ stated primary funders showed this was incorrect in many cases. A further 24% said they “didn’t know”, so that nearly 60% lacked good knowledge of their funders’ requirements.
The influence of funders on journal selection, however, appears to be growing. Although this was the least important factor in journal selection, it was nonetheless described as important or very important by 18% of respondents, up from 15% in 2014. Funder influence was highest in Asia (judged as important or very important by 27% of Asian repondents), and lowest in N America and among HSS researchers (both 14%).
The factors used by authors to select journals for submission have been studied before, and there were few surprises here. The most important factors cited were journal reputation, Impact Factor, relevance, quality of peer review, and speed of peer review, with STM researchers placing more emphasis on Impact Factor, and HSS researchers on journal readership. On the other hand, having the option for immediate open access, offering cascade peer review, and funder influence were of little importance to most researchers.
One of the advantages of this survey was its exploration of what researchers actually meant by journal reputation. Most important seems to be the perception of where their peers choose to publish their best work. Encouragingly, consistency of quality was also important; my feeling is that many researchers see consistency as a reflection of good peer review and active editors.
The Impact Factor retains its dominance as a key measure of journal quality, particularly among “next generation” researchers. It was the second most important element in assessing journal reputation, and ranked 1st, 2nd or 3rd by over 60% of respondents. Researchers in Medicine placed the most weight on it, and those in HSS the least. Its importance was greatest among researchers from Asia, followed by Europe, and then N America. Younger researchers also attached more weight to it than older ones.
Readership is an important factor in journal selection, but what aspects of it do researchers value? For journal reputation, what matters is the size of readership within the researcher’s own discipline; conversely, the geographic spread of readership, and having readership among interdisciplinary and industrial or government audiences, are all relatively unimportant. Similarly, in describing their ideal audience, researchers have (on average) a strong interest in being read by their peers, with only limited interest in researchers from outside their own field or practitioners in their field, and have very little interest in everyone else (including lay audiences or policy-makers).
Finally, what publisher services do researchers value? Those highly valued were the core activities: constructive peer review; making the article discoverable; and rapid peer review. On the other hand, least valued were the arguably peripheral services such as supporting rich media; promoting the paper; offering cascade peer review; and pre-submission services. It’s possible that some of these score low on average but remain important to a minority of authors, or in particular circumstances; it may also be that things like supporting rich media are simply taken for granted these days.
Mark is a publishing consultant to the STM publishing and information sectors. Prior to establishing his own consulting practice in 2003, he held positions at Ingenta, CMP Information, and IOP Publishing. He is the author of The STM Report: An overview of scientific and scholarly journal publishing. More at: www.markwareconsulting.com