Designing for Expansive, Sociotechnical Futures: Conjecture Maps as Boundary Objects Between Learning Scientists and Computer Scientists

• In Event: The Role of Emergent Technologies in Equitable Futures: Navigating Between Entrenched Injustices and Expansive Possibilities

Abstract

Within the education community, speculative design methodologies have risen to the forefront with the recognition that many of education’s foundational issues— specifically equity issues—cannot be addressed through incremental design interventions (Gutierrez & Jurow, 2016; Garcia & Mirra, 2023). This recent shift has coincided with major advances in emergent technologies (e.g., large language models) that some have envisioned to revolutionize education (Kasneci et al., 2023). Consequently, computer scientists (who previously may not have had a social orientation to their work) have increasingly become key drivers in the design of learning interventions, and have also brought in their disciplinary-specific practices and commitments (Zhang et al., 2021). In this project, we focus on one such disciplinary practice that often adversely constrains the speculative: the practice of solution mapping (Lee et al., 2020), i.e., starting from currently feasible technical capabilities and imagining futures through this lens. In this project, we demonstrate how practices structured around a well-recognized tool in the learning sciences—conjecture maps—can be used to specifically support interdisciplinary teams of computer scientists and learning scientists in seeing beyond dominant computational imaginations—towards expansive possibilities for education.
Conjecture mapping (CM) is a tool for visualizing connections between learning designs, learning theory, and desired learning outcomes (Sandoval, 2013). CMs have now been widely used for proposing and managing research projects on learning technology topics (Wilkerson, 2017; Ahn et al., 2021). Towards our goal of manifesting speculative possibilities, we make two key extensions to existing components of CMs: learning outcomes and design embodiments. First, we explicitly include as a learning outcome speculative narratives and the desired goals of speculative spaces—possibilities initially unencumbered by technical feasibility. Second, we expand design embodiments to include back-end design, i.e., elements of design that computer scientists are directly concerned with, but may not be directly visible to an user.
In this paper, we study how these modified CMs were used within the Institute for Student-AI Teaming, a team of learning scientists and AI researchers building an AI partner that supports small group collaborations. Prior to technical developmental work, Institute researchers ran a Learning Futures Workshop, which engaged non-dominant youth in imagining expansive possibilities for academic collaboration and AI. Following this workshop, Institute learning scientists and computer scientists worked closely together to be responsive to youth’s speculative hopes and concerns.
Our claim about modified CMs is that they effectively scaffold researchers in reversing solution mapping, i.e., forefronting critical imaginations and then creating tight connections to the technical designs that computer scientists are most familiar with. To investigate, we reviewed extensive data over a year, including: field observations, surveys to Institute researchers, and past versions of CMs. We use a methodology of instrumental case studies (Stake, 1995) to study facilitator moves that accompanied CM activities and supported interdisciplinary teams in being responsive to the speculative possibilities proposed by youth. Researchers found CMs to be a powerful tool to both align speculative orientations across a team and to forefront the integration of learning theory into the technical design.

Authors

• Michael Alan Chang, Boston University

• Jeremy Roschelle, Digital Promise Global

• Rachel F. Dickler-Mann, Denver Zoo

• Jeffrey Bush, University of Colorado - Boulder