A New Context for Professional Networks: Understanding the Social Structure of #NGSSChat Through Social Network Analysis

• In Event: Educators Meet the Fifth Estate: Social Media in Education

Abstract

Objectives
In science education, the #NGSSchat community on the social media platform Twitter has grown into an active community centered on discussion, through regularly-occurring ‘chats’, of the new standards and associated changes in teaching and learning. Despite the prominence of #NGSSchat, little attention has been paid to its use by scholars and the wider education policy community. We aim to understand whether and how #NGSSchat is serving as a professional network, and how #NGSSchat may be distinct in the context of other, similar social media-based communities (i.e., #commoncore; Supovitz, Daly, del Fresno, & Kolouch, 2017). We focused on two key network-related processes: selection (Spillane et al., 2012) and influence (Frank et al., 2004). While much of the past research on professional networks in education have focused on in-person sources of data, social media use has emerged as an alternative source of information for educational SNA research (Gruzd & Haythornthwaite, 2013; Hu, Torphy, Opperman, Jansen, & Lo, 2018), as in this study. In particular, we ask:

1. Who (in terms of occupation) has participated in #NGSSchat?
2. What explains greater interactions between #NGSSchat participants?
3. What characteristics are associated with sustaining an individual's participation in #NGSSchat?

Method
In this study, we make use of digital traces of individuals’ (n = 248) interactions from archived social media posts from one academic year (2014-2015) of activity (n = 7,400). We processed the Twitter content into relational data focused around two kinds of interactions, conversations (replying or mentioning) and endorsing (favoriting, retweeting, or quote Tweeting) in order to understand who has participated in #NGSSchat and where they were located through qualitative coding of profile data. We specified a multi-level p2 selection model that we estimated via the brms R package (Buerkner, 2019) to understand why some interactions occur more frequently than others through the hashtag. We also used a Poisson generalized linear model in order to understand how interactions influenced individuals to sustain their participation in the subsequent academic year (2015-2016).

Findings
Findings revealed that individuals who have participated in the #NGSS hashtag hold a variety of science education roles from a variety of locations in the United States, with teachers being the most frequently identified group, followed by researchers. In terms of selection-related process, group membership matters more in terms of the conversations that for endorsements (with teachers being highly-conversed with but not highly-endorsed), but that being actively involved was associated with a high degree of interaction. Being highly-involved is not the only consideration that is important, as who interacts with one is also important: receiving conversations from individuals who were central or active in the network was most strongly predictive of sustained involvement.

Significance
In all, this work suggests that #NGSSchat involves conversations among stakeholders, rather than simply conversation and interaction from researchers and administrators to teachers and that participation in this informal community can be sustained over time. This work provides a basis for subsequent work not only on #NGSSchat and the social network analysis of other social media-based professional networks.

Author

• Joshua Michael Rosenberg, The University of Tennessee - Knoxville