The following proposal, Modules for Statistics Graduate Teaching Assistants Learning to Teach Equitably with Authentic Data (GTAs-LEAD), is submitted for consideration as a Level 1, Track 1: Engaged Student Learning proposal seeking to develop, implement, and research evidenced-based professional development modules for statistics graduate teaching assistants (GTAs). Developing data acumen is necessary for every citizen to harness the data revolution in their workplaces and everyday lives (NASEM, 2018). There have been longstanding recommendations to reform introductory statistics courses to address the growing need for expanding opportunities to investigate authentic data (GAISE College Report ASA Revision Committee, 2016; Ridgway, 2016). Despite the numerous efforts to develop curricular resources to bring authentic data investigation into introductory statistics classrooms, the instruction often focuses on procedural aspects of statistical skills. Moreover, in large universities, such as North Carolina State University (NCSU) and Michigan State University (MSU), the responsibility of leading these active learning opportunities often falls on graduate teaching assistants (GTAs) who may not have any formal education in teaching (Justice et al., 2017). Without discipline-specific professional development for statistics GTAs, innovative curricular resources are unlikely to reach the fruition of equitable learning outcomes of developing data acumen. The GTAs-LEAD project will address this urgent need to facilitate teacher learning of statistics GTAs. The project team hypothesizes that with carefully organized discipline-specific teacher learning modules, statistics GTAs can learn to teach equitably with authentic data while working with their GTA communities. Guided by a design and development research approach, the GTAs-LEAD project will: (1) design a set of four research-informed modules for statistics GTAs learning to teach equitably with authentic data (LEAD Modules), (2) implement LEAD Modules with two GTA communities teaching introductory statistics courses at NCSU and MSU, and (3) further refine LEAD Modules based on design-based research that examines GTA development and their communities. By drawing on the interdisciplinary expertise of the PI team, the GTAs-LEAD project infuses knowledge bases and resources from statistics education and mathematics teacher education to support statistics GTAs learning to teach equitably with authentic data.

In this Level I early stage Design and Development study on Teaching, using design-based research, we aim to design, implement, investigate, and iteratively refine a video-based coaching model to develop mathematics teachers��� responsive pedagogies for linguistically marginalized students. Building from cutting edge research on linguistically responsive mathematics pedagogies (Adler, 2021; de Araujo & Smith, 2021; Marshall et al., in press; Song & Coppersmith, 2020), this project addresses the persistent need to foster mathematics teachers��� learning about supporting linguistically marginalized students (Lucas & Villegas, 2010; Prediger, 2019). Our novel model centers the experiences of students through video clips as rich tools for teacher learning. Our approach builds from key findings from our small prior study: that video-based coaching can support teachers in learning justice-oriented pedagogies such as social justice mathematics (Marshall, 2022) and learning to disrupt racialized patterns of exclusion in mathematics classrooms (Marshall, 2020) by supporting teachers��� sensemaking about their own students��� unique experiences in mathematics classrooms and giving timely, formative feedback as teachers encounter problems of practice (Horn et al., 2022). Central to our model is this core insight: that classroom video holds potential for supporting teacher learning of responsive pedagogies because of its opening of a window into students��� experiences, proximity to practice, context-embeddedness, and affordances for troubleshooting such pedagogies soon after teachers try them in their classrooms. The scholars collaborating to lead this project have strong histories of work designing and investigating professional development for educational equity, and complementary expertise to build a powerful and scalable model for mathematics teachers learning of responsive pedagogies. Our overarching research question is: How do secondary mathematics teachers learn about supporting linguistically marginalized students? The primary outcomes of this research include: a portrait of the challenges and opportunities that mathematics teachers face in supporting linguistically marginalized students, an iteratively refined model of professional development for teachers��� learning of responsive pedagogies, and an empirically-grounded theory of teachers��� learning to support linguistically marginalized students.

With data science rapidly growing as a practice and profession, it is necessary to build a model for data science education that can engage, support, and empower students from diverse programs of study. It also has been widely reported that students from underrepresented communities often face systemic marginalization while developing STEM identities. To address these issues, the NC State Data Science Academy (DSA) has developed and piloted the All-campus Data science through Accessible Project-based Teaching and learning (ADAPT) model. This instructional model was designed to encourage students with diverse disciplinary and sociopolitical identities to participate in data science education through project-based learning with purposeful choices that reflect students��� various identities. ADAPT courses are taught by a diverse team of instructors who have a variety of experiences utilizing data science in multiple settings (e.g., government, academia, industry). The project team hypothesizes that offering learning experiences that (a) view students��� unique identities as strengths and (b) empower them to make choices in how they do data science can help students see themselves as data scientists and foster a sense of belonging in data science communities. Guided by a design-based research paradigm, this project will test and refine the ADAPT model while developing instructional resources for broad implementations within and potentially beyond NC State.