Research on teachers’ use of social media has typically assumed that it is a) driven by a need for professional learning and b) best understood in terms of individual motivations. In this study, we use a dataset of nearly 600,000 tweets posted to one or more of 48 Regional Educational Twitter Hashtags associated with 44 U.S. states. To explore the influence of local contextual factors on hashtag- and account-level activity in these hashtags, we use an analytic approach heretofore uncommon in social media-focussed education research: generalised linear and multilevel modelling. At the hashtag level, higher numbers of teachers within a state, proportions of students receiving subsidised meals, student-to-teacher ratios, and amounts of state spending per child are associated with more activity within a regional hashtag; by contrast, more left-leaning state governments and citizenries are associated with less activity. At the account level, more experienced accounts and accounts in more right-leaning states contribute more tweets to these hashtags. These findings reinforce established understandings of Twitter as a site for teacher learning; however, they also underline the importance of acknowledging other important purposes of teachers’ Twitter use, including receiving emotional support and engaging in activism.

Twitter hashtags may serve as valuable means for teachers’ professional development. However, given the diversity of hashtag spaces and teacher needs, teachers must assess a given hashtag and compare it to their learning needs and preferences before determining whether it would be helpful. To support this reflection, I examine data associated with 60 Regional Educational Twitter Hashtags (RETHs) during the first six months of 2016 to begin describing the variety of teacher learning-focused Twitter spaces and make distinctions between them. My results indicate that these RETHs vary according to their relative focus on sharing, intimacy of personal connection, and volume of activity, each of which has implications for professional development. The dimensions resulting from this study may prove helpful for teachers, teacher educators, and hashtag coordinators.

The existing work on teacher-focused Twitter hashtags typically frames each hashtag as a single, unified phenomenon, thereby collapsing or erasing differences between them (and any resulting implications for learning). In this study, we conceived of teacher-focused hashtags as affinity spaces potentially containing subspaces distinguished by synchronous chats and other, asynchronous communication. We used computational methods to explore how participation differed in terms of content, interactions, and portals between these contexts within the #michED hashtag used by Michigan teachers. During the 2015–2016 academic year, #michED saw more non-chat activity than chat activity, and most participants only engaged in one mode of activity or the other. Participation during chats was associated with more replying as well as more socially-, affectively-, and cognitively-related content, suggesting a focus on social interaction. In contrast, non-chat participation was associated with more retweeting, mentioning, hyperlinks, and hashtags, suggesting a focus on content dissemination. These results suggest that different affinity spaces—and different literacy practices—may exist within the same hashtag to support different objectives. Teachers, teacher educators, and researchers should therefore be careful to make these distinctions when considering Twitter as a learning technology for teachers.

Although researchers have discovered a great deal about who uses Twitter for educational purposes, what they post about, when they post and why they participate, there has so far been little work to explore where participants in educational Twitter contexts are located. In this paper, we establish a methodological foundation that can support the exploration of geographical issues in educational Twitter research. We surveyed 46 participants in one educational Twitter hashtag, #michED, to determine where they lived; we then compared these responses to results from three digital methods for geolocating Twitter users (human coding, machine coding and GPS coding) to explore these methods’ affordances and constraints. Human coding of Twitter profiles allowed us to analyze more participants with higher levels of accuracy but also has disadvantages compared to other digital—and traditional—methods. We discuss the additional insights obtained through geolocating #michED participants as well as considerations for using geolocation and other digital methods in educational research.

Affinity spaces are digital or physical spaces in which participants interact with one another around content of shared interest and through a common portal (or platform). Among teachers, some of the largest affinity spaces may be those organized around hashtags on Twitter: These spaces are public, largely unmoderated, and thriving, yet very little is known about them, especially those based in geographical areas such as American states. This paper examines these potential affinity spaces by providing the first large-scale study of them in the form of an examination of 47 State Educational Twitter Hashtags (SETHs). Collecting over 550,000 tweets over 6 months, our analysis focused on who is participating in SETHs, how active participants are, and when participation occurred. We found support for two of Gee’s tenets of affinity spaces, in particular many interactions through a shared portal. Though the content of tweets were not the focus, this study’s findings lend support to efforts to identify which particular SETHs will be best suited to subsequent analysis of their content and what times subsequent analysis might most productively focus on. We discuss implications for how we conceive of teacher professional development and suggest directions for future research focused on the content of tweets associated with SETHs.