Currently, I am conducting a broad analysis of contemporary mobile learning research. I was looking forward to this exercise because I wanted to contextualize the results of our concept mapping study with ongoing research. In this post I list a few high-level views on the mobile learning research based on statistics reported by Thompson's ISI Web of Science.  A side objective of my research is to identify specific developments in Africa and South America, so don't be surprised to find references to this region.

I was rather excited to have direct access to Thompson's ISI Web of Science - of which I thought only as a citation counter until this point. This minor misconception was created by the lack of access to the query interfaces back at the OUNL. I was excited to find the different ways to make use of relatively abstract concepts of the citation indices for dwelling into mobile learning research slightly differently as usual.

I started with a simple query on the "mobile learning" topic for the smart phone era (that is everything that was published since 2005) and I got 13K+ hits from all kinds of topic domains. So I triggered a filter to get only the journal articles on the topic. Of course this excludes the Journal on Mobile and Blended Learning (IJMBL), but still I got more than 1000 hits across all topics. After excluding topic areas that were way too hardware related such as electronics and electronic engineering, and topics that use a different understanding of learning such as artificial intelligence or autonomous systems, I got around 800 hits covering all SCI and SSCI journals. Around 20% of these papers were still off topic and mostly related to wireless networking hardware or artificial intelligence. 

The first finding was that about 80 journal papers were published alone in 2011 and the early issues of 2012. With the peak of mobile learning in 2010, with 100 individual publications on mobile learning in SCI or SSCI indexed journals it shows that almost half of all top rated publications were published in the last two years. Another interesting detail is that in 2011 and 2012 BJET, Computers & Education,  IEEE TLT and JCAL published about a half of these papers, while Computers & Education was the clear lead with 15 papers alone - which is about the same as a single volume of IJMBL.

The second finding was that among the most cited papers Computers & Education is the single most relevant journal counting for one third within the top 100 most cited mobile learning related papers. Within this ranking, BJET, Computers & Education, ET&S, and JCAL count for 80% of the most cited papers (again only counting SCI & SSCI journal papers). In order to get into this ranking you need to have at least 5 references from other papers within the ISI WOS index, which would be slightly above the average citation of the papers found with my filter.

So I ventured a tiny bit into the origin of the publications by looking at the authors' names. Of course this is not a valid way of doing such an analysis, but since I know a few names in the field you get a feeling from which part of the world the papers were submitted. Interestingly, the most frequently submitting and cited region was east asia, while no papers from Africa were present in both charts. This is confirmed by the statistics of the most publishing authors. Among the top ten authors in the field is one canadian (Kinshuk) and the remaining authors are of Asian origin. The single most productive author is Gwo-Jen Hwang from Taiwan with 17 SCI/SSCI publications since 2005.

By running an analyzis on the overall contributions by country and region it showed that the USA was most active with 226 papers. However, there were hardly any recent (2011/2012) or "relevant" (in terms of frequently cited) submissions found in the rankings during the same period. Again the African hemisphere is represented with less than 20 publications since 2005. 

Without looking at the contents of any paper this analysis shows trends and gaps of international research on mobile learning. As I only included journal articles in this analysis, the picture might slightly change for conference proceedings. However, I won't expect any major shifts in the results.