While no participants identified their primary affiliation as a policymaker or government representative, 7% of participants (n = 5) defined their second stakeholder category as policy/government. This study was approved by the university research ethics

board at the University of British Columbia and all participants provided informed consent. The first step of the concept mapping method included a brainstorming session to generate the initial statements or ideas. At a time and place of convenience, participants accessed a web-based platform (Enterprise Feedback Management; Vovici Corporation, Herndon, VA) to participate in this initial asynchronous task. Participants completed the five demographic questions then responded find more to a single question or focal prompt. The foreword statement and focal prompt for participants PFT�� manufacturer included: “There may be many aspects of the built environment (i.e., sidewalks, street connectivity, etc.) and the social environment (i.e., community connectedness, social supports, etc.) that impact older adults’ outdoor walking. These could include aspects that promote or limit walking. “From your perspective, aspects of the built

environment and social environment that influence older adults’ outdoor walking are We refined the scope and wording of our focal prompt after pilot testing with our project team; and concluded that the prompt resulted in responses that were either facilitators or barriers to outdoor walking. In the full protocol, we did not limit the number of responses participants could contribute to process. Three authors HH, CS, MA synthesized the responses in preparation for sorting and rating tasks; this included breaking down complex responses into their component parts, and clarifying the language used to ensure understanding across

stakeholder groups. We removed duplicate statements, or statements reflecting very similar content. The second step of the concept mapping method is sorting and rating of the brainstormed statements. The core stakeholder group completed the sorting and rating tasks using the Concept Systems Global software (Concept Systems, Inc., Ketanserin Ithaca, NY). Participants electronically sorted synthesized statements into groups they perceived to conceptually relate; they could create as many groups as best represented statements. We asked participants to rate each statement on two constructs, importance and feasibility to implement; on a scale from 1 (low) to 5 (high) and scored relative to the other statements. After sorting and rating, we used the Concept Systems Core software to analyze data using multidimensional scaling and hierarchical cluster analysis.