Beyond Problem Solving: Framing and Problem-Solution Co-Evolution in Data Visualization Design

Abstract: Visualization design is often described as the process of solving a well-defined problem by navigating a design space. While existing visualization design models have provided valuable structure and guidance, they tend to foreground technical problem-solving and underemphasize the interpretive, judgment-based aspects of design. In contrast, research in other design disciplines has emphasized the importance of framing--how designers define and redefine what the problem is--and the co-evolution of problem and solution spaces through reflective practice. These dimensions remain underexplored in visualization research, particularly from the perspective of expert practitioners. This paper investigates how visualization designers frame problems and navigate the dynamic interplay between problem understanding and solution development. We conducted a mixed-methods study with 11 expert practitioners using design challenges, diary entries, and semi-structured interviews. Through reflexive thematic analysis, we identified key strategies that participants used to frame problems, reframe them in response to evolving constraints or insights, and build bridges between problem and solution spaces. These included using metaphors, heuristics, sketching, primary generators, and reflective evaluation of failed or incomplete ideas. Our findings contribute an empirically grounded account of visualization design as a reflective, co-evolutionary practice, where framing is not a preliminary step but a continuous activity embedded in design. Participants often reshaped their understanding of the problem based on solution attempts, tool feedback, and ethical or narrative concerns. These insights extend current visualization design models and highlight the need for frameworks that better account for framing and interpretive judgment. (See paper for full abstract.)