Last week (Oct 7 – 9) the altmetrics community made its way to Amsterdam for 2:AM (the second altmetrics conference) and altmetrics15 (the 4th altmetrics workshop). The conference is aimed more at practitioners while the workshop has a bit more research focus. I enjoyed the events from both a content (I’m biased as a co-organizer) as well as logistics perspective (I could bike from home). This was the five year anniversary of the altmetrics manifesto so it was a great opportunity to reflect on the status of the community. Plus the conference organizers brought cake!
This was the first time that all of the authors were in the same room together and we got a chance to share some of our thoughts. The video is here if you want to hear us pontificate:
From my perspective, I think you can summarize the past years in two bullet points:
- Amazing what the community has done: multiple startups on altmetrics, big companies having altmetric products, many articles and other research objects having altmetric scores, a small but vibrant research community is alive
- It would be great to focus more on altmetrics to improve the research process rather than just their potential use in research evaluation.
Beyond the reflection on the community itself, I took three themes from the conference:
More & different data please
An interesting aspect is that most studies and implementations rely on social media data (twitter, mendeley, Facebook, blogs, etc). As an aside, it’s worth noting you can do amazing things with this data in a very short amount of time…
However, there is increasing interest in having data from other sources or having more contextualized data.
There were several good examples. Juan Pablo Alperin gave a good talk about trying to get data behind who tweets about scientific articles. I’m excited to see how better population data can help us have. The folks at altimetric.com are starting to provide data that looks at how articles are being used in public policy documents. Finally, moving beyond articles, Peter van Besselaar looking at data derived from grant review processes to study, for example, gender bias.
It’s also good to see developments such as the DOI Event Tracker that makes the aggregation of altmetrics data easier. This is hopefully just the start and we will see a continued expansion of the variety of data available for studies.
The role of theory
There was quite a bit of discussions about the appropriateness of the use of altmetrics for different tasks ranging from the development of global evaluation measures to their role in understanding the science system. There was a long discussion of the quality of altmetrics data in particular the transparency of how aggregator’s integrate and provide data.
A number of presenters discussed the need for theory in trying to interpret altmetrics signal. Cameron Neylon gave an excellent talk about his view of the need for a different theoretical view. There was also a break out session at the workshop discussing the role of theory and I look forward to the ether pad becoming something more well defined. Peter van Bessellaar and I also tried to argue for a question driven approach when using altmetrics.
Finally, I enjoyed the work of Stefanie Haustein, Timothy Bowman, and Rodrigo Costas on interpreting the meaning of altmetrics. This is definitely a must read.
Going beyond research evaluation
I had a number of good conversations with people about the desire to do something that moves beyond the focus of research evaluation. In all honesty, being able or tell stories with a variety of metrics is probably why altmetrics has gained traction.
However, I think a world in which understanding the signals produced by the research system can be used to improve research is the exciting bit. There were some hints of this. In particular, I was compelled by the work of Kristi Holmes on using measures to improve translational medicine at northwestern.