Technological interventions for the development of social skills among autistic children have shown promise. However, the majority of the tools have been developed from a high-resource and Western perspective and may not be culturally or materially appropriate for a low-resource environment. While the use of participatory and co-design approaches is strongly advocated for autism-related research, the direct involvement of autistic children is not always feasible due to the presence of social stigma and the constraints of the environment. In such cases, therapists and teachers are often engaged as proxies, yet the methodological implications of proxy-based co-design have not been sufficiently explored. In response to such shortcomings, this paper presents a process‑oriented methodological analysis of the co‑design of POMA (Picture‑to‑Object Mapping Activities), a tangible user interface–based iPad application developed to support the social skills of autistic children in Sri Lanka, with therapists and teachers acting as proxies. Through ten rounds of iterative workshops, we present the evolution of co-design methods through breakdowns and negotiations across app critique, paper prototyping, wireframing evaluation, and two iterations of prototypes. Our analysis reveals methodological developments from (i) tensions in app critique from a Western perspective based on linguistic and cultural differences, (ii) resource constraints on participation, (iii) tensions in multidisciplinary co-design leading to design divergence and consolidation, (iv) material failures in environmental sensing leading to redesigns in tangible sensing, and (v) participant turnover leading to repeated onboarding and alignment. The findings highlight that proxy-based co-design in low-resource, culturally specific contexts is a dynamic and contingent process, requiring ongoing methodological adaptation rather than the straightforward application of established participatory practices. By focusing on challenges and adaptations, this paper contributes empirically grounded methodological insights into proxy-based co-design for autism technologies, with implications for researchers working in resource-constrained and culturally diverse settings.
Soysa et al. (Fri,) studied this question.