Photovoice: Goals, current research, and gaps.

Photovoice is a process whereby community members take photographs and tell stories about their communities, engage in dialogue, and present their work through exhibition to policymakers [23]. We view photovoice as the enactment of hyperlocal identification of risk, vulnerability, and potential for action. Photographs and stories from participants focus on their direct experience in their communities and represent concerns for the locations that impact their daily lives. As such, photovoice is a central component of our novel community-based participatory approach to stakeholder engagement and climate adaptation planning.

A distinct extension of photo novella [24] that was formally introduced by Wang and Burris in 1997 [24], this qualitative method of inquiry empowers individuals to capture experiences in their communities, promote dialogue that produces collective knowledge, and communicate with policymakers and other stakeholders in positions of power [25]. Due to its flexibility and adaptability, photovoice is particularly well suited for community-based participatory research [23, 26]. Over the last two decades, photovoice has grown in popularity [26, 27] and been used in contexts from physical and mental health [27, 28] to community development [29] and been shown to be accessible to people in a range of age groups and other demographics [30, 31]. Photovoice can be delivered effectively in person or online [32, 33].

An emerging area of interest in photovoice research relates to the environment, emergency management, and climate resilience planning [34]. In this context, investigators have solicited the experiences and perspectives of members of rural and indigenous communities recovering from severe rain and wind events [35], coastal fishing village residents experiencing climate-driven marine ecosystem changes [36], and women in Nepal facing mental health impacts from drought, water scarcity, and other climate related changes [28]. In work in a coastal setting in Australia, Chandler and Baldwin [37] described the use of photovoice as participants shared stories demonstrating community understandings of climate change effects and their capacity as a persuasive tool. Many of these studies have sought to understand the role of social capital in community resilience, and to explore its place in equitable climate governance [38].

In addition to co-producing new knowledge and developing potential solutions to local problems, the process of photovoice can build strength and resiliency within individuals and communities [23], while increasing social capital [38]. Additionally, the photovoice approach is responsive to the concerns, such as those expressed by members from Isle de Jean Charles, Louisiana, threatened by sea level rise, regarding the need for, and importance of, community-led approaches and solutions [39]. As such, photovoice is “highly consistent” with the core principles of community-based participatory research, “stressing empowerment and an emphasis on individual and community strengths, co-learning, community capacity building, and balancing research and action” [40]. However, photovoice research also has gaps, limitations, and unintended consequences.

While photovoice can empower communities and individuals, the prioritization of crosscutting community themes and dialogue [35] is often secondary to the individual stories and exhibitions, critical thinking during photovoice could be strengthened [41], and the process of generating solutions and advocating for collective priorities is seldom reported. Photovoice also often lacks external supporting data related to the topic of investigation (e.g., climate, health). While the goal of individual experience as essential to the creation of knowledge cannot be overstated, empowerment and improved decision-making may also benefit from additional education about the topic, and access to relevant data and the governmental departments generating and analyzing that same data. Indeed, photovoice rarely includes geospatial data as part of the process [34].

Derr and Simons [25], in their review of photovoice in environmental and conservation contexts, report inconsistent engagement with decision- or policymakers. This inconsistent engagement and potential lack of follow-through may lead to unintended consequences. For example, Pritzker, LaChapelle, and Tatum [31] found that although photovoice empowered youth, those same youth also expressed frustration at the perceived hypocrisy of adults around civic engagement. These findings highlight the need for ongoing discussions among community organizations and decision- and policymakers to enable the reinforcement of community narratives and implementation of desired solutions.

While photovoice is a powerful tool of place-based engagement, Russo et al. [35] further highlight a lack of clarity in how to best capitalize on its strengths, address its limitations, as well as the need for, and process of, engaging in comparative analysis across communities, a process that is further hindered by the lack of details provided about coding schemes and analytic techniques [34].

These limitations can diminish the effectiveness of photovoice for community action and advocacy. By viewing photovoice as a designed protocol of interaction [19–21] we anticipated and addressed the limitations described above with the integration of a design thinking component, which advances actionable potential through its framework for prioritization and problem-solving processes, an education module on climate change risk and geospatial mapping (including access to generate their own maps), and the inclusion of other community organizations and policymakers throughout the process to ensure meaningful and ongoing engagement.

Indeed, we see the lack of consistent engagement between community and policymakers in community-based approaches as a missed opportunity of stakeholder engagement. While photovoice research focuses on community participants as the primary stakeholders, we would suggest that the role of policymakers as stakeholders in the photovoice process should be reimagined. Rather than being brought in at the end of the process, engagement with the community throughout could strengthen policymaker commitment and buy-in to community-based approaches. Engaging with community organizations and policymakers from the beginning of the process allows all participants to interact over longer periods of time, rather than the one-shot interactions that occur in typical photovoice exhibitions. We see this as a way to help build relationships and commitment to community- engaged processes by policymakers (not just participants), letting them see how participants construct and talk about issues over a period of time. This also gives extended voice to community members with opportunities to engage at multiple times with policymakers.

Design thinking: Goals, current research, and gaps.

The second element of our process, design thinking, provides a human-centered framework for participants to explore circumstances, conditions and opportunities, determine key challenges, imagine new opportunities to solve problems and reach ideal conditions, and build out a particular solution through facilitated activities. Associated initially with David Kelley, founder of the firm IDEO, and Stanford University’s Hasso Plattner Institute of Design (the d.school), design thinking has been widely applied in systems and product design as in the Cocreator Lab at Philips and IBM Design at IBM, and in business operations and products, as in the development of MassMutual’s Society of Grownups [42]. Liedtka describes design thinking as a social technology through which participants “collaborate and agree on what is essential to the outcome” [43, p. 79]. The premise of design thinking is that the process can be applied to any condition, and it begins with a close identification with the people most impacted by the outcome. The key steps of design thinking advance from an understanding of human needs and interests to the testing and prototyping of ideas in relation to their impact and effort. More recently, design thinking has been shown to promote critical thinking [44] and to be effective in bridging the gap between individuals and policymakers in conditions that lie outside conventional problem definitions [45].

The Stanford d.school design thinking process outlines five steps: Empathize, Define, Ideate, Prototype, and Test (https://www.alnap.org/help-library/an-introduction-to-design-thinking-process-guide), and is typically conducted in relation to a focus area selected by the organizers of the session. Empathy mapping builds from observation and engagement with users or the affected group. From these observations, team members define leading challenges and engage in ideation, a free-flowing process of imagining possibilities. Coalescing around a prioritized idea, a solution is developed as a storyboard or mockup, and finally tested across expansively defined users. Philips Design built on this method to develop their Cocreator Lab based on its "innovation and you" philosophy, seeking to establish a broader application of design thinking to organizations and collaborative problem solving.

In comparing process improvement to a design thinking orientation, Roberts et al. [46] noted that “process improvement prioritizes evaluation of limited set of possible solutions,” while design thinking “prioritizes comprehensive understanding of underlying problems.” They view this a distinct advantage in addressing “problems that have unpredictable solutions (wicked problems)” as design thinking encourages divergent thinking.” Instead of focusing on what is significant to individuals within a defined context, design thinking seeks to discover what matters in daily life to a group. Mintrom and Luetjens [45] analyze the potential for design thinking as an added method for “policymakers to create interventions and services that improve the user experience and enhance public value.” They believe that the successful application of design thinking depends on a “diversity of skills and abilities,” as well as “curiosity and openness.”

Design thinking is used in multiple disciplines and for multiple problems. Academic literature on design thinking for climate adaptation often has a strong focus on the design of the built environment [47–49] or on specific policy or industry related innovation, such as Govaerts and colleagues [50] report on design thinking for Agri-food systems that are threatened by climate change. These processes tend to be top-down processes, led by experts or leaders focused on key problems, and largely engaging designers or experts. There is little published literature focusing on design thinking approaches to climate adaptation driven by community members.

Viewing design thinking as an engineered process of communication and engagement [19–21] allows for the modification and integration of the design thinking process with photovoice to help create a stronger approach to community-based climate adaptation approaches. Where photovoice encourages the discussion of themes within photographs, photovoice studies are often lacking in any collective approaches to problem solving, relying instead on individual exhibition of stories and discussion with policy makers [41]. Design thinking offers a powerful complement to photovoice that encourages critical thinking [44], focuses on prioritizing issues and themes across participants, and enables a unique approach to problem solving. Although the design thinking process is often conducted via a top-down perspective, where policymakers gather participants to address a specific issue (see, for instance, 50), the HyLo method seeks to provide an alternative model that asks participants from the local community to bring their individual and collective issues (drawn from their photovoice stories) while gathering policymakers to engage with their process. In this way, the HyLo method retains the organic, grassroots nature of participant photovoice stories while providing a framework for communities to define and advance their collective agendas. Building on the themes residents identified through photovoice as the basis for design thinking, the direct experiences of community members determine the scope of topics engaged. The primacy and immediacy of the individual photovoice narratives inform the design thinking process to provide a community driven process of lived experience. The HyLo team’s involvement in the design thinking process focuses on the facilitation of ideas while the participants bring the content, priorities, and possible solutions. The presentation of photovoice stories through an initial exhibition lends both context and personal impact to the design thinking process.

Synthesized together, the design thinking processes of engagement, dialogue, and problem solving further support the underlying goals of photovoice of improving individual and community capacity, building social capital, and increasing agency. Additionally, the introduction of geospatial mapping at various points in the process provides community members access to data on social and climate assets and risks that many policymakers rely on to make decisions. The HyLo team has developed this mapping process into an Integrated Climate Risk Assessment (ICRA) tool which consists of multiple data layers with demographic, geologic, and infrastructure data for each neighborhood enabling access to a body of locally relevant information to guide in community-based policy planning. Finally, the inclusion of representatives from community organizations and policymakers throughout the entire process enables community member interaction over multiple sessions.

Participants.

In total, 28 community members participated in some, or all, of the sessions, including 10 from Little River and surrounding communities, and 18 from Homestead. Of those who responded to our demographic questionnaire, our participants were majority female (n = 11), male(n = 2); reported diverse race/ethnicity—Black or African America (n = 6), White (n = 3), Hispanic (n = 9), other (n = 3); reported education levels as high school or GED (n = 1), some college (n = 4), college or technical degree (n = 4), master’s degree or higher (n = 3). Participants ranged in age from 21–54, with a mean age of 34 years. Twelve participants responded that English was their first language while one responded that Spanish was their first language. Several participants indicated second language proficiencies, including Spanish (n = 5), English (n = 1), and Haitian Kreyol (1). Participants were split between renting and boarding rooms (n = 5) and home ownership (n = 6) with 2 participants not knowing or disclosing their housing status. Participants reported living in their community between 1 month and 18 years, with a mean of 8.3 years. The results presented below are derived from the 22 participants who participated in three or more sessions.

Overall, the demographics of the participants are fairly representative of the communities at large in terms of race/ethnicity, age, and home ownership versus rental status. Participants in our study reported higher levels of education and included a higher percentage of women than the roughly 50 percent ratio reported in both communities.

Sessions.

Covid-19 forced a transition from in-person to online engagement. Sessions were conducted in collaboration with our community partners, The CLEO Institute and Catalyst Miami. The University of Miami HyLo team provided technical support by hosting Zoom meetings, assigning breakout rooms, collating photovoice stories, preparing, annotating and capturing design thinking slides, and assisting in facilitation where needed. Community partners led the sessions and served as primary facilitators.

With the shift to online workshops, we realized that interaction would be limited if all participants met in one large group. As such, we designed sessions around breakout rooms that were limited to four to six participants each. For the Little River sessions, we kept all six community members together who attended beyond the first session; for Homestead, we opened each session with all members together and then moved into four breakout rooms with four participants each. Additionally, members from city and county government, including the City of Miami’s Departments of Planning & Zoning and Public Works, and Miami Dade County’s Offices of Resilience, and Miami-Dade Department of Parks, Recreation and Open Spaces observed each session and engaged in discussion, where appropriate.

We opened with a joint session to introduce climate change concepts and the photovoice process to all participants. We held three subsequent sessions for each location—three for Little River and three for Homestead. We kept the groups separate for these subsequent rounds given their different locations and to enhance the potential for participants to form a sense of shared community and focus on local needs. Sessions included photovoice story presentations and discussion of common themes, an introduction to maps of local conditions relevant to photovoice concerns, and an introduction to design thinking, with a focus on discovery of experiences and frames for understanding. A second session of design thinking, through Discover, Framing, and Imagine (ideation and an impact-effort matrix to prioritize potential solutions) advanced to the closing phase of Build with relevant partners and actors identified. Discussion with key local and government leaders concluded the workshops.

Photovoice coding and analysis.

While we provide illustrative examples of photovoice stories and narratives by community, coding and analysis was designed to provide different data than would normally be reported in a photovoice study, and to address concerns about transparency in coding reported in the literature [35]. Rather than focusing specifically on the narratives, our objective with this study is to compare key concerns across communities–and as such, systematic coding provides a mechanism for comparison [35]. To address specific concerns related to each community, we moved beyond the broad themes generated by participants in the sessions to develop a comprehensive set of primary and secondary codes (e.g., storms as primary, with wind damage, evacuation, and rainfall as secondary) related to climate change and adaptation. Photovoice stories were coded based on the most precise code for the story, and stories could have multiple codes. Additionally, photovoice stories were coded for their valence: positive (opportunity), negative (threat), or mixed (threat and opportunity). For each photovoice story, we coded the written narratives by first identifying the most prominent theme, then coded for each additional theme present in the narrative. To determine nuances in the stories themselves, we engaged in closer analysis that examined the patterns of stories and how primary and secondary themes wove meaning that is similar or different across communities. To do this, we examined each story for all codes to determine patterns within themes and subthemes.

Design thinking data and coding.

Data for the analysis of design thinking comes from the annotated design thinking worksheets each group developed as they moved through the process. There is no current systematic method for analyzing design thinking data and processes. As such, we focus on the evolution of narrative and ideas as they progress through the process from Discover to Build. We let the themes and progressions speak for themselves through the broad presentation of data as it reflects both the logics and lived experiences participants incorporate in developing solutions to local problems.

Measures for individual and community capacity.

To determine the impact of our process on community and individual capacity we created or adapted existing measures to assess community assets and risks and to evaluate the effectiveness and impact of our community engagement process in increasing: social connectedness, community engagement, individual agency, threat assessment, communication competence, and communication apprehension.

Social Connectedness measures the degree to which a person feels meaningful connection within their neighborhood [53] and was measured using an adapted 5-item Likert-type scale (strongly disagree to strongly agree) with reliability of Cronbach’s alpha = .87. Items include statements such as “There is someone in my neighborhood I could talk to about important decisions in my life”, and “I have relationships in my neighborhood where my know-how and ability are recognized.”

Individual Agency items were adapted from Shealy, Godwin & Gardner [54] and measured using a 5-item Likert-type scale with reliability of Cronbach’s alpha = .87. Items included such statements as “I can make my neighborhood stronger to deal with future climate change,” and “I can be a resource for my neighborhood for knowledge about future climate change and what we can do about it.”

Communication Apprehension was measured using four items from McCroskey [55], Personal Report of Communication Apprehension, and presented a reliability of Cronbach’s alpha = .55. Item analysis resulted in dropping two items which improved reliability to .71. Items included statements such as “Generally, I am comfortable while participating in meetings/group discussions about climate change,” and “I like to get involved in group discussions about climate change.”

Communication Competence was measured using three items from McCroskey’s communication competence measure [55] and had a reliability of Cronbach’s alpha = .90. Participants responded to the statement “Please indicate how good you would be at talking about climate change and what to do about it” for the following items: Talk in a group or meeting of friends or acquaintances; Talk in a group or meeting of strangers; Talk in a group or meeting with members of local government.

Threat Assessment items were adapted from “Florida: Public opinion on climate change” [56] and split into perceived threat and threat likelihood. Perceived threat from climate change was measured using two-items on a 5-point Likert type scale from 1 –not at all serious to 5 –very serious and had a reliability of Cronbach’s alpha = .92. Items included: How serious of a threat is climate change to you and your family? and How serious of a threat is climate change to your local neighborhood? Threat likelihood was measured using a 6-item Likert type measure on a 5-point scale from 1 –not at all likely to 5 –very likely and had a reliability of Cronbach’s alpha = .89. Items included statements assessing the likelihood of neighborhood impacts from stronger and more frequent storms and hurricanes, droughts and water shortage, flooding, and sea level rise. Finally, the perceived timing of risk was measured on a 5-point Likert type scale ranging from 1 –now to 5–100 years from now.

Neighborhood Asset Quality was measured using 10 items and had a reliability of Cronbach’s alpha = .98. Participants ranked what they liked (from strongly dislike to strongly like) about their neighborhood on items such as: housing (affordability and type), quality of local services, access to public transportation, and parks and open spaces.