This is a repository with analytic code and instructions on how to request the data for: Rosenberg, J. M., Reid, J., Dyer, E., Koehler, M. J., Fischer, C., & McKenna, T. J. (in press). Idle chatter or compelling conversation? The potential of the social media-based #NGSSchat for supporting science education reform efforts. *Journal of Research in Science Teaching*. ## Abstract > The Next Generation Science Standards (NGSS) chat (#NGSSchat) is a social media-based professional network used to discuss topics related to the NGSS in the United States. While successful reforms involve and coordinate the work of multiple stakeholders, recent research points out a striking lack of coordination between the individuals working in different educational roles—to the detriment of intended changes in the system. In this study, we analyzed more than 7,000 posts from individuals participating in #NGSSchat on Twitter (n = 247) during two years of one-hour synchronous discussions. We studied the depth and types of conversations that took place, the extent to which the involvement of teachers, administrators, researchers, and organizations was balanced, and what explains participation in the network over time. Using a mixed-methods approach involving social network analysis, we found that conversations were primarily transactional, or social, and substantive, or providing opportunities for sense-making about the standards and for participants to transform their practice and that individuals from diverse roles participated, with teachers comprising the plurality of those involved. Additionally, researchers, administrators, and teachers were the most active in the network, with no differences in both initiating, or sending, and being the recipients of, or receiving, replies as a part of conversations. Finally, we found that being a teacher or administrator, as well as receiving replies from individuals who were important in the network, were positively related to sustained participation in the network in the following year. We discuss how #NGSSchat—as a social media-based professional network—demonstrates similar features in other effective networks, and how social media-based networks invite new visions for how to implement ambitious, large- scale changes in science education. ## Analytic code We used R (specifically R Markdown documents) to carry out the analyses. The R packages used across all of the analyses are loaded at the beginning of each of the following four files: - Code to create the sociograms (network visualizations) are in: `descriptives.Rmd` - Code for the frequency of different types of conversation used in the analysis of data for research question 12 are in `conversations-1.Rmd` - Code for the selection models used in the analysis of data for research question #2 are in `balanced-participation-rq2.Rmd` - Code for the influence models used in the analysis of data for research question #3 are in `sustained-involvement-rq3.Rmd` ## Accessing original data on #NGSSChat A unique aspect of our dataset is that we used data that #NGSSChat moderators had self-archived tweets via a service for doing so, Storify (which no longer exists). We accessed these via a data mining approach, and then prepared them for use in this study using the [{rtweet}](https://docs.ropensci.org/rtweet/) R package, which provided a great deal of additional information on each tweet. Because of the provenance of these tweets, being self-archived by #NGSSchat moderators - and that they are no longer available through the Storify service - we wish to share this data with others who are interested in accessing or using it. This is also important to us for purposes of transparency, in the spirit of open science. At the same time, we wish to protect the privacy of #NGSSchat participants whose tweets may appear in this dataset. So, to balance these two values (sharing the data openly while also taking steps to protect #NGSSchat participants' reasonable expectations of privacy), to access the original data, we have implemented a tiered access system requiring an application and discussion with the authors. To apply to access the data, please complete this form, and message Dr. Joshua Rosenberg (jmrosenberg@utk.edu) to indicate that you have submitted it: https://forms.gle/THXQY6uZ526vA12p7