Reframing is key to mitigating the risks of implicit and inaccurate assumptions when dealing with complex, open-ended problems. While behavioural designers regularly face such problems, reframing is overlooked in current behavioural design guidance. Therefore, there is a need to better understand and demonstrate reframing's potential impact in behavioural design. We address this need via an exploratory, controlled experiment with design-engineering students responding to a complex, open-ended problem with a significant behavioural component. We evaluate the impact of three reframing stimuli against a control, measured with respect to behavioural design quality. The three stimuli included a structure-only stimulus (sequential steps of actions), a content-only stimulus (unordered prompting questions), and a combined structure-content stimulus. To evaluate behavioural design quality, we conduct a mixed-methods assessment of design outputs at different points in the design task: ideation of possible problem-solution perspectives, mindmapping, and proposition of a final solution (concept). Our findings confirm that all three stimuli are effective in increasing behavioural design quality, with increased emphasis on behavioural aspects and enhanced integration of behavioural and technical aspects of problems and solutions. This contributes to understanding the importance of reframing in developing problem-solution understanding in behavioural design, with significant implications for theory and practice.

The increasingly transdisciplinary context of design, where designers collaborate with other disciplinary and domain experts, means there is a growing need to evidence the effectiveness of design methods. We address this need in two ways. First, we propose a ‘chain of evidence’, from motivation to claims, operationalising this in a systematic assessment framework. Second, we systematically review current design method research. Our results reveal that while all links in the chain of evidence are reported across the literature and best practices can be identified, no individual paper either reports all links or consistently achieves best practice. Our framework and results demonstrate the need for standards of evidence in this area, with implications for design method research, development, education, and practice.