Last December we compared the geographic distribution of HASTAC user base and commenters and confirmed that language was a major factor governing the geographic distribution of users, with commenters based mostly in English speaking locations and areas where English is spoken as a second language. The user base, conversely, is geographically more diverse with hotspots in Europe, Australia, and regions of Africa -- areas with very low counts in the commenters map.
In the previous months we also explored the relationship between the geographic distribution of HASTAC Scholars and areas of disciplines of interest. We found that connections were prevalent in the northwest, south, and east coast, particularly Texas and California. The results made it very difficult to figure out which topics were covered by any of the clusters we plotted in the map, although we managed to identify major topics like digital humanities, social media, literature, and media studies, but also pedagogy, technology, gaming, cultural studies, and history.
This time we mapped the geographic distribution of user-assigned tags and site-proscribed topics by U.S. State (from 2006 to 2013). The map above shows the four most common tags assigned by users to their profiles (HASTAC groups). This is only possible because HASTAC user profiles incudes groups that are topics-oriented, and a small portion of HASTAC users have contributed geographic information. The animated map below shows the frequency of topics by U.S. state based on this source of information. The dynamic map includes 30 topics ranked from the most to the least frequent. If the topic you are interested is not there, you can see all 63 plots of topics we managed to map here (again ranked from the most to the least frequent).
There are a number of caveats to these maps. First of all, just over 10% of topics could be geographically mapped because very few users provided location and topics to their profiles (the actual proportion is 13%). The distribution is pretty skewed and most tags come from few users located in Texas, North Carolina, and California that provided geographic information and assigned tags to their profiles. Another drawback of this sample is that there is little variation across topics because most data points are coming from just a few nodes.
This sample is considerably small compared to the number of topics and tags assigned to posts by users. Although it takes a somewhat convoluted process to match blog posts to geographic locations, we manage to retrieve the location of just under half of the blog posts in the period. Unfortunately, as you can see here, the proportion of tags assigned by users based in North Carolina is so much higher than average that it eclipses any activity elsewhere. One possible solution is to exclude tags coming from North Carolina and highlight the overall distribution of topics over the remaining states, as shown in the map below.
Despite the potential distortion introduced by sample size discrepancies, the maps shed light on the topics and areas that are relevant to HASTAC users. Given the skewed distribution of the data and the limited sample size, these maps should be viewed as preliminary and largely exploratory rather than conclusive. As always, please leave us a comment if you find anything unusually interesting (or just plain unusual) about these maps.
This material is based upon work supported by the National Science Foundation under Grant Number 1243622. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.