Co-Designing Data Dashboards From Research-Practice Partnerships and Learning Analytics Perspectives

• In Event: Using a System of Practical Measures, Routines, and Representations to Support Instructional Improvement at Scale

Abstract

Objectives: In this paper, we highlight how the design of data visualization and dashboard tools develops from direct co-design with research-practice partnerships (RPPs). The mere presentation of data in charts, graphs, and other visualizations is not a trivial or purely mechanical process. Instead, we present several case study vignettes, where RPPs and data visualization methods are combined to derive creative new dashboard tools that flow directly from the work of partners and hold promise to support local processes.

Theoretical Framework: Most work in the Learning Analytics and Data Visualization fields focus on utilizing theories of perception or cognition to create charts, graphs, and other representations of data (Alhadad, 2018; Klein et al. 2006; Lee et al., 2016). Many evaluations of these data visualizations focus on whether users perceive and understand the data, but researchers have observed a disconnect in designing dashboards that truly support educators’ work processes (Few, 2006; Jivet et al.2017). In addition, methods from RPPs offer ways to co-design solutions with partners, ensuring that newly designed tools flow directly from local work processes (Author, 2019). We illuminate how combining these design approaches and strategies help us to identify key tensions to design for, translate the knowledge of partners into design ideas, and develop data tools that are potentially more usable and relevant to educators as they embark on improvement work.

Methods: We developed design case studies using data from meeting notes of the RPP teams and the design team, interviews with educators, and field notes and artifacts from co-design sessions. The interviews with partner teachers and coaches involved presenting sample classroom data and inviting them to think aloud and provide feedback on how data could be used for improvement practice. We documented and tracked the evolution of our designs of tools through an open coding process and illustrate how information from our RPPs translate into key design decisions.

Results: We undertook design work to develop dashboards that instructional coaches could utilize to aid in their support of teachers. Three cases illuminate how we leveraged deeper understanding of the RPP teams, coaches’ work processes, and concerns to develop visualizations of data. We will present examples of data dashboards tuned to teachers’ work, the nuanced tweaks made when developing dashboards for instructional coaches, and finally how issues of privacy and trust that are commonplace in school settings and collaborative work with data become infused into the design of dashboard tools. Taken together, we aim to illuminate design principles for creating tools that are attuned to the system of improvement work that partners undertake, and move beyond current notions of human-centered and learning analytics design.

Significance: Critical questions remain about how one might design data visualization and dashboard tools that flow more directly from the needs of educator partners, and facilitate local work processes more readily. Our case vignettes provide strategies and methods for this process, building on and synthesizing approaches from learning analytics, data visualization, and RPPs to contribute new approaches for design partnership research.

Authors

• June Ahn, University of California, Irvine

• Fabio Campos, New York University

• Ha Nguyen, University of California - Irvine

• Maria Hays, University of Washington-Seattle Campus

• Jan Morrison, Vanderbilt University