Data Science in Educational Research
Data science is commonly characterized as a field that integrates disciplinary expertise, statistics, and advanced computational techniques to answer questions or solve problems. In education, data science education has potential as a set of ideas and practices that cut across STEM content areas and can also be meaningful to learners, but the potential for data in education has not yet been fully realized. My research focuses on how learners think of and with data, particularly in science education settings. I refer to this research interest as in data science education research. I have pursued this research interest through:
- developing new pedagogical approaches to support students to participate in aspects of data science that are critical but under-emphasized, such as structuring data and combining data from different sources (e.g., Rosenberg, Edwards, & Chen, under review);
- exploring how students engage in foundational aspects of data science, especially recognizing, modeling, and visualizing variability in data (e.g., Rosenberg, 2018; Rosenberg, Schmidt, & Koehler, under review);
- using computational and machine learning methods in a way that is better aligned with the goals of educational researchers (Krist & Rosenberg, 2016; Rosenberg & Krist, under review), including work recently funded through the National Science Foundation (NSF; NSF Grant No. 1920796)
- and, adapting existing pedagogical approaches and technological tools, such as science simulations, to output complex, quantitative data for students to model (Rosenberg & Lawson, 2019).
Data Science Education
New approaches to research such as data science allow researchers to study teaching and learning in innovative ways, but, most of the time, these approaches are used to augment existing approaches, instead of transforming how research is carried out. My second research interest concerns how scholars can leverage data science-related methods in order to study teaching, learning, and educational systems. I refer to this research interest as in data science in educational research. I have pursued this research especially to study learning in the context in which it takes place—and to do so at a finer-grain size than past methods have permitted. I have pursued this research interest through:
- making use of a novel experience sampling method to understand how students and youth experience learning science in-the-moment and in authentic classroom contexts (e.g., Beymer, Rosenberg, & Schmidt, 2018; Schmidt, Rosenberg, & Beymer, 2018), including work recently funded by the NSF (NSF Grant No. 1937700);
- advancing new research approaches, such as social network analysis to analyze teachers’ conversations on social media (Rosenberg, Greenhalgh, Wolf, & Koehler, 2017; Rosenberg et al., under review);
- making existing statistical tools easier to use through meta-programming techniques (e.g., Heiss, Anderson, & Rosenberg, 2019; Rosenberg, Beymer, Anderson, & Schmidt, 2018);
- and, building the capacity of educational researchers and other education stakeholders through writing about data science and computational social science research methodologies and carrying out workshops on the use of R (e.g., Bovee, Estrellado, Mostipak, Rosenberg, & Velásquez, under contract; Dai, Rosenberg, & Lawson, forthcoming; Rosenberg, 2019; Rosenberg, Lawson, Anderson, Rutherford, & Jones, forthcoming).
Open Science and Building the Capacity of the Field
Data science is separate from open science, but there is an overlap between them, in that data science techniques facilitate sharing work—at all stages—in a more open manner. I have an interest in open science and have the commitment to carrying out scholarship in a way that is more open to those interested in it. As an example of this has to do with the book, Data Science in Education Using R (Bovee, Estrellado, Mostipak, Rosenberg, & Velásquez, under contract). While this book is to be published by Routledge in 2020, we asked the editor with whom we are working if we would be able to share a website-only version of the book. Our editor noted that such an agreement had not been made by Routledge before, but, after exploring the topic, included in our contract that we will retain the copyright (and permission to publish) a website-only version of the book (while Routledge will own the copyright for the print and e-book versions). This will make the book more accessible to a wider audience of educators and education stakeholders interested in the topic.