You’re an early-career higher education researcher? One of those hybrid academics who doesn’t really fit into traditional typologies? Don’t lose hope, not yet. In most of my studies, I was tempted to fall towards computational techniques and reporting methods, or to automate cumbersome tasks as much as possible. Those powerful temptations normally won the battle. To make it clearer, I’m a geeky sociologist – someone who’s never accepted as a pure sociologist among other sociologists but never gets accepted as a pure coder among computer scientists, web developers, or programmers either!
Luckily for me, there’s an increasing interest towards “computational social science” (see this conference series, for example), or pretty much towards anything with the adjective computational in front of it. Although these types of skills might sound fancy for some, they might come in handy when you want to communicate your results in a form that is different from traditional black and white, long, and hard-to-read articles. Let me give you an example: One of my first tasks at DZHW (German Centre for Higher Education Research and Science Studies) was to prepare a world-wide view to Higher Education Research and Science Studies (HERSS) field with a specific focus on how Germany has been active in this world. It was supposed to look into publications data from Web of Science (WOS) or Scopus and complement it with other available information. This is how the final product came to look like:
The dashboard (click to see) aims to explore the world of Higher Education Research and Science Studies (HERSS) by providing an overview of the key actors in the field. We also focus on the research centers and organizations in Germany and their local and international collaborators. To create the dashboard, we queried all the publications in 23 most prominent journals in the field (see the journals section in the list below). By looking at the cities, country of affiliation, higher education institutes, and organizations, we tried to see which countries and organizations occupy the key positions. We extracted all publications (29,898 ranging from 1971 to 2017) from Clarivate Analytics’ Web of Science (from the data infrastructure of the German Competence Centre for Bibliometrics) that are published in one of the 23 main journals in HERSS. The list of these journals was extracted from website of the German Federal Ministry of Education and Research (BMBF). The data included the titles of articles, keywords, abstracts, publication years, the names and affiliations of the authors, and number of citations received.
The Dashboard sections explained
- HERSS World
- HERSS World Map: presents publications, historical information, and the main collaborators in HERSS field for each country. It also shows the main organizations that have been collaborating with Germany in HERSS research. It presents the first and last HERSS publication year, importance of each institution in the collaboration network based on normalized degree centrality scaled to 0-1 with 1 showing the highest influence and it provides the list of institutions collaborating in HERSS field. Choosing Past 10 years on the map presents the same information with a specific focus on 10 most recent years (from 2007 –2017).
- Continents/Countries: presents the distribution of HERSS organizations in each continent, country, and city which are sized based on the count of HERSS publications.
- German Cities: presents the distribution of HERSS organizations in German cities which are sized based on the count of HERSS publications.
- Journals: presents the number of publications in each of the 23 main HERSS journals and the earliest and latest HERSS publication years.
- Organizations: presents all HERSS organizations, the earliest and latest HERSS publication years and total number of HERSS publications. In a separate tab it shows the organizations active in the past 10 years.
Why did I implement HERSS in such a way?
Before starting with this task, I had some discussions with my supervisors and colleagues on how they wanted the results to look like. They were kind to give me a good deal of freedom to choose how I would like to communicate the outputs. I chose to use “Flexdashboards” – one of the newest gifts the amazing people at Rstudio has created for the R community. It gives you the possibility to integrate your R code and results and present it either in static (like my example) or dynamic form with more interactivity, which is called Shiny applications. The next version of my dashboard is going to be like that (see other examples here).
If you want to obtain suitable computational skills that allow you to compete in academia and keep up with the new trends, I have a few suggestions: First, develop a powerful toolbox (consisting of any of the available choices, such as R, Python, Stata, Julia, to name a few) and try to master it as well as possible. Second, try to get your head around all aspects of data science: the idea of asking questions, gathering suitable data to answer those questions, cleaning the data, modelling it, and finally interpreting the data and reporting the results in the most attractive way. Finally, don’t limit yourself to one or two of the above skills. Go beyond the traditional ways of communicating research results. After all, if you want your research to have an impact, you need people to remember it! With the right set of tools, both you and your results will remain in the long-term memory of your audience.
[This post is an edited repost of “Key Actors in Higher Education Research and Science Studies (HERSS)”, posted on December 4th, 2018].
Aliakbar Akbaritabar (short: Ali) has studied sociology in different academic contexts for 14 years. Since September he has been a postdoctoral researcher at DZHW (German Center for Higher Education Research and Science Studies). His research focuses on Research System and Science Dynamics.