The number and variety of online tools and platforms for all phases of the research cycle has grown tremendously over the years. We have been charting this ‘supply side’ of the scholarly communication landscape, first in our figure of 101 innovative tools in six different phases of the research cycle (Fig 1.), and subsequently in our growing database of tools for 30 distinct research activities within these phases.
To get a visual impression of the development of tools over time, we plotted the 600 tools currently in our database against the year they were created (Fig 2., also available on Plot.ly).
New online tool development rose sharply at the end of the 1990s and again and the end of the 2000s. The recent rise of 2013 and 2014 may be an artefact of the way we have been collecting these tools since 2013: through (social) media mentions, reviews in journals and crowdsourcing. All three sources focus on tools that have just been launched. Apart from special circumstances in higher education and research there may also be effects here of the dot.com bubble at the end of the 1990s and the web 2.0 explosion in the second half of the 2000s. The interesting peak of new outreach tools in 2008 is lacking a clear explanation. The slump in 2015 for all types of tools is due to the fact that it easily takes 6-12 months before new tools attract media attention.
A more detailed view of tool development emerges when tools are plotted separately for the different activities within research phases, against the year (and month, where possible) they were created (Fig 3, also available on plot.ly). As an extra layer of information, we added the current number of Twitter followers (where available) as a proxy for the interest a tool has generated.
In interpreting this plot, there are some important considerations regarding the underlying data that should be taken into account:
- We have limited inclusion to online tools specifically, excluding tools that are available as download only. Also, browser extensions are not included.
- The picture for the early years (up to c. 2006) is less complete than that for more recent years. For example, early traditional publisher platforms are not included and tools may have risen and fallen during this period and thus not have been added to our list.
- We have assigned each tool to only one research activity. This affects the picture for tools that can be used for multiple research activities (like Mendeley, Figshare and ResearchGate).
- The number of tools for the publication phase are distributed over many separate research activities, resulting in a seemingly less densely populated plot area.
- It takes a while for (new) tools to accrue Twitter followers, which is why tools created in 2015 have relatively few Twitter followers.
- The number of Twitter followers is one of many possible measures of interest/popularity, each with their own limitations (see for example, this blogpost by Bob Muenchen). Our global survey on research tool usage that is currently running, will provide more substantiated data on actual tool usage.
Taking these considerations into account, some interesting observations can be made:
- The slow start of the development of online tools for academia is remarkable, given that the internet was developed (at universities !) decades ago. The first website was launched in 1991 and the first graphical browser (MOSAIC) was introduced two years later, but it took until 1997 before PubMed became available and another four years before the first web-based reference management tool (RefWorks) was launched.
- Of the 30 research activities we identify, search (nr. 3), experiment (nr. 9), publish (nr. 24) and outreach (n.25) have had the longest continuous period of (online) tool development. For these activities, many online tools have been developed prior to 2008. Activities for which tools have only become available more recently are getting access (nr. 4) and post-publication peer review and commenting (nr. 27/28).
- Relatively few tools exist for sharing posters and presentations specifically, and no new ones have been developed recently. However, there are many other tools (like FigShare and ScienceOpen) that enable archiving posters as one of their functionalities.
- Tools for the writing, outreach en assessment phases often have many followers, perhaps because these tools are often relevant for all research disciplines (and, for writing and outreach, even beyond academia). Tools for discovery and publication are more often discipline-specific, and might reflect persistent differences in publication cultures and the desire for selectivity.