The search resulted in 2773 records after removing duplicates or records in other languages or in the wrong format. A further 2688 records against the inclusion criteria were removed through title and abstract screening. Of the remaining 85 reports, 6 reports were not retrieved, and 66 against the inclusion criteria were then excluded, leaving a set of 13 papers (Fig. 1).
The integral interventions in the reviewed papers can be categorised into three types. The first type focuses on technology learning that is supported by the physical and social environment, for example, a community formed by OA to learn technology from each other at a designated physical meeting point (36). The second type focuses on social interaction that is supported by the physical environment and technology, for instance, a community for sharing, learning and managing music using technology where people can also meet at a designated physical location (37). The third type focuses on physical activity that is supported by the social environment and technology, for example, a community-based walking programme supported by digital technology (38). The CBPR principles have been followed to various extents by the reviewed studies, while no studies have covered or reported on the last two CBPR principles (Fig. 2). Twenty-four design considerations were generated in 8 categories based on the eight CBPR principles (Table 3). Details of the intervention types, categories, design considerations and their associated example quotes can be found in Additional file 4. Each subsection below will elaborate on one category of design considerations.
Table 3
Categories | Design considerations |
Community | Recruit participants1 via the commonly used channels that are specific to the target community |
Understand the values and norms of the community |
Enable connections among stakeholders2 of the community |
Resources | Utilise physical resources of the community |
Utilise social resources of the community |
Understand the history and politics of resource allocation within the community |
Collaboration | Co-design the intervention with key stakeholders |
Match the level of co-design based on the shared goals of stakeholders |
Help stakeholders to envision future scenarios |
Mutual benefits | Help participants concretise the benefits of the intervention in their daily lives |
Uncover reasons behind tension among stakeholders (if any) |
Clarify whose interests the stakeholders represent |
Co-learning | Scaffold education and training based on the current knowledge, preferences, and experiences of participants |
Cultivate a collaborative learning culture in the community |
Facilitate learning between communities with similar interests |
Flexibility | Start small and set milestones to track the progress |
Collect feedback and ideas from key stakeholders regularly and iterate the intervention when needed |
Understand the reasons for dropping out |
Inclusion | Uncover and tailor to the diversity of participants |
Recognise any ageism views held by the stakeholders and researchers themselves |
Know what technologies the participants are familiar with (have both technical and non-technical options for participants if needed) |
Dissemination | Provide periodical findings to stakeholders |
Create a practical guide to help communities implement the intervention |
Train the responsible person(s) for intervention handover in sufficient time |
(1with participants, we mean people who will participate in the intervention; in our case, they are OA. 2with stakeholders, we mean not only participants but also individuals involved in organising, maintaining, and creating conducive conditions for the intervention.)
Community
In the context of established communities, researchers need to recognise and respect the unique identity of each community. The first design consideration is to recruit participants via the commonly used channels that are specific to the target community. It has been found that each community has its own channels for information dissemination and communication. For example, Yang & John (2020) approached a senior centre to recruit participants (39), and den Haan et al. (2021) recruited participants via a local association and local newspaper (40). Frei et al. (2019) tried diverse channels, including local events for OA, local newspapers, flyers to the mailboxes of OA, and a pharmacy poster (38). Community members usually stay up-to-date and trust the information disseminated from these channels, which can lead to higher levels of interest and participation.
The second design consideration is to understand the values and norms of the community. Shared values and norms play an important role in the psychological sense of community (41). Most action research studies conducted observations, interviews or focus groups to understand what the members of the community think are important in life and the things that are regarded as usual by them (36–39, 42–44). For example, researchers learned from their observations and interviews that “social connectedness” and “physical health” are very important to OA in the senior centre, and Bingo is the most popular game they play. Hence, they developed an intervention called Team Bingo, which aims to “reduce social isolation and increase physical activity” for OA in a community setting (39).
The third design consideration is to enable connections among stakeholders of the community. Many studies have explored different ways to connect key stakeholders of the community in their interventions (38, 39, 42, 43). For instance, Frei et al. (2019) established a WhatsApp group to facilitate communication among participants and organisers, where scheduling walking activities emerged as a popular topic (38). Gooch et al. (2021) aimed to enhance participants’ sense of connectedness through a community display (42). Keirnan et al. (2019) organised a co-design workshop leading to intervention concepts for improving communication between OA and community managers (43). O’Brien et al. (2021) highlighted that their intervention, focused on tracking physical activities, received feedback from participants emphasising the importance of social connections for program enjoyment and engagement (45). Lazar et al. (2021) reported that even though OA were invited to co-design a makerspace, many OA reported that they did not want a makerspace in the first place; they missed communication with managers regarding their needs (44).
Resources
This category focuses on how to make use of the resources of the community during intervention development. The first design consideration is to utilise the physical resources of the community, i.e., resources in the physical environment of the community. For example, Frei et al. (2019) made use of existing walking trails in the neighbourhood (38). Gooch et al. (2021) conducted workshop at a location where community members usually meet (42). Lenstra (2017) found that OA socialised more in senior centres than in public libraries when they learned technology and received support on technology use (36), which indicates how the social atmosphere can be shaped by the physical environment.
In parallel to the first design consideration, the second design consideration is to utilise the social resources of the community. We define community social resources as the value derived from member relationships, shared knowledge, exchanged skills, and members’ potential ability to obtain resources, favours, or information from their personal connections. For example, Lazar et al. (2021) observed that during the formation of the makerspace, community members with technological expertise were selected for the committee, and beyond the committee, some members supported this initiative through donations and sharing their expertise (44). Gooch et al. (2021) encouraged OA to organise walks by themselves and recruit new members via their connections (42). In the smartphone learning program, OA previously trained in using technology offered classes to other OA in the community (40). In the Sharing Dance program, facilitators were identified locally for each site (46).
The third design consideration is to understand the history and politics of resource allocation within the community. Understanding how resources have been allocated in the past and who makes or influences decisions is important for successful resource utilisation. For instance, Pedell et al. (2021) explored resources, enablers, and barriers to social prescribing services in the community (47). Lenstra (2017) highlighted ageist policy in public libraries, where volunteers for technical support are predominantly recruited from universities, mainly in their 20s (36). This practice perpetuates the notion that OA should seek help from younger generations rather than peers for technology assistance. To counteract ageist policies, OA organised themselves in a community to pressure the local government for financial support to form new senior centres for technology learning. To understand local politics, some researchers have also investigated how decision-makers are perceived by community members (43, 44).
Collaboration
This category focuses on how to ensure that key stakeholders work together to develop the intervention. The first design consideration is to co-design the intervention with key stakeholders. Involving key stakeholders actively and early in the development process can generate creative ideas, facilitate consensus, and promote collaboration. Many studies worked with OA collaboratively, and many of them involved other stakeholders during co-design, such as community workers (42), staff in the senior centre (39), staff in social prescribing services (47), and key persons from local policy and service organisations (38). Kosurko et al. (2022) involved community administrators, family carers, volunteers, and staff facilitators in addition to OA (46). Keirnan et al. (2019) posit that co-design “breaks down the power structures between managers and residents” (43).
The second design consideration is to match the level of co-design based on the shared goals of stakeholders. Three studies started the co-design based on a concept that researchers already drafted, for example, walking activity (38), social prescribing (47), and community displays (42). The stakeholders were involved to co-design the details of these concepts; three studies started without preconceived concepts from the researchers, used co-design to understand the challenges of the community, and interventions were then developed (39, 47, 48); one study helped the community to realise its own initiative, where the researchers positioned themselves as facilitators (45). As illustrated by the studies above, three levels of co-design emerged in our review. When the level of co-design did not match the shared goals, resistance and scepticism from under-involved groups might occur (44).
The third design consideration is to help stakeholders envision future scenarios. Gooch et al. (2021) used storyboards to help OA and community workers understand how a community display could be used for encouraging physical activities (42), and participants were given templates to create storyboards by themselves. Keirnan et al. (2019) used LEGO to encourage residents of the retirement park to build an ideal park in their minds, based on which the researchers identified the wishes of the residents and the current challenges in the park (43). Yang & John (2020) developed a prototype of their intervention and presented it to OA and staff at the senior centre for a test play to help them understand what roles the intervention can play in the future (39).
Mutual benefits
This category focuses on ensuring that all stakeholders gain benefits from the intervention. The first design consideration is to help participants concretise the benefits of the intervention in their daily lives. This is the most common for studies focusing on promoting physical activities, as some regard exercise as a pain, and it is hard for some participants to keep envisioning the long-term gain of physical exercise. O’Brien et al. (2021) meet participants regularly to explain the benefits of physical activities in their daily lives (45). Frei et al. (2019) invited each OA to set short- and long-term goals that they wanted to achieve by participating in the intervention (38). Mao et al. (2020) found that letting OA realise their accumulated competencies by participating in the intervention over time has positive consequences for their self-efficacy and confidence in realising their goals (37).
The second design consideration is to uncover reasons behind tension among stakeholders, if any. By understanding the reasons behind this, it is easier to resolve the conflict by finding a common ground; communicating the interests explicitly could also help stakeholders understand each other. For instance, Keirnan et al. (2019) organised a codesign workshop to let residents of a retirement park express their tensioned relationships with community managers using arts and crafts materials (43). The residents later reported that they had more understanding about the managers, “it is not them who are problematic, it is the system that they are in”. The study on the formation of the makerspace discovered that different stakeholders have varied opinions on what was considered a suitable return on this investment, which partly led to resistance to the intervention (44).
The third design consideration is to clarify whose interests the stakeholders represent. When talking about “OA”, participants might think of people who are older than them, and they then “design for others”, rather than “design for themselves”. It is also valuable when people design for others, yet it is important to clarify if this is the case. For example, when briefing the community workers during the development of the community display, they were prompted to discuss “what they felt would be useful for the OA they support”, rather than how they would support OA with the community display (42). The same happens when talking about “community”, as Lazar et al. (2021) put it, “there is a complexity to whom participants believed constitutes this community” (44). Clarifying questions can help researchers understand who is speaking for whom and then determine if these imagined others exist and their interests were predicted correctly.
Co-learning
This category intends to help researchers provide education and training that empowers participants in the intervention while also learning from the participants. The first design consideration is to scaffold education and training based on the current knowledge, preferences, and experiences of participants. Especially when participants need to learn the technology in the intervention. For instance, in the intervention using trackers to promote physical activities, the researchers inquired about the previous experiences of participants with tracking devices and then provided tailored support for each participant (45). While developing the makerspace, it was observed that OA got stuck or gave suggestions not suitable for a makerspace, i.e., some OA do not know what a makerspace is (44). This indicates that more scaffolding is needed at the beginning of the co-design sessions. Regarding the preferences of learning, some OA think that learning something new is exciting already, while others only learn if they could benefit loved ones or others (44).
The second design consideration is to cultivate a collaborative learning culture in the community, i.e., community members can provide education and training to each other. The study on community music found that enabling competence sharing could support social participation of OA (37). OA in the community radio intervention also expressed that they enjoy sharing and learning experiences with and from each other (49). Lenstra (2017) noted that newcomers to senior centres have to be aware that they are expected to participate in technology support services when becoming part of the group, and OA only felt comfortable helping each other when an officially designated “technology helper” was in the room, whom they could turn to if obstacles arose. Therefore, specific criteria need to be met for this collaborative learning culture to develop.
The third design consideration is to facilitate learning between communities with similar interests. For instance, Reuter & Liddle (2020) hosted a festival to connect groups of OA who create radio content with each other (49). This interconnectedness allows communities to learn from one another, leading to valuable cross-pollination of knowledge and experiences. When establishing the makerspace, Lazar et al. (2021) showcased various makerspace examples from other communities to inspire and engage the participants (44). By exposing the community members to successful models from elsewhere, the project leader sought to generate interest and ignite creativity among participants. Thus, existing examples from other communities provide valuable lessons and ideas that can be adapted and applied to the specific context of the targeted community.
Flexibility
This category aims to support researchers in developing interventions gradually with feedback cycles. The first design consideration is to start small and set milestones to track the progress of the development. “Starting small” implies both a lower complexity of the intervention and a smaller number of participants. For instance, Gooch et al. (2021) conducted a pilot study to evaluate the storyboards they created, and these storyboards were then revised ready for the workshop (42). The workshop collected feedback from stakeholders about the concept before it was realised into a working prototype. Kosurko et al. (2022) conducted six pilot studies of the Shared Dance Older Adult program, and the feedback from each pilot resulted in the evolution of a multimodal program (46). Milestones can help researchers and stakeholders keep track of and align on the progress of the project. The study of Frei et al. (2019) serves as a good example: intervention design, goal setting with participants, monthly reflection meetings, responsibility transfer, researcher departure, and follow-up (38).
The second design consideration is to collect feedback and ideas regularly from key stakeholders and iterate the intervention when needed. Being regular helps to build a routine and manage the expectations of key stakeholders. For example, Frei et al. (2019) set a monthly reflection meeting during intervention deployment (38). When designing the social prescribing service, Pedell et al. (2021) organised three workshops for the insights to cascade into the next to ensure that the co-design process was “open and flexible” (47). Lazar et al. (2021) observed that the committee met monthly to decide on many aspects of the makerspace (44). Yang & John (2020) involved both staff and OA from the senior centre to ensure that the design suits the needs of both groups (39).
The third design consideration is to understand the reasons for dropping out. Two studies conducted exit interviews to collect this information, which could reveal useful feedback for intervention improvement. For instance, Frei et al. (2019) understood from their exit interviews that the participants dropped out because of adverse events unrelated to the intervention (38), while O’Brien et al. (2021) learned from their exit interviews that participants drop out due to either the usability of the tracking device or lack of motivation towards the intervention (45).
Inclusion
This category intends to help researchers ensure that the intervention has considered a wide range of physical, cognitive, socioeconomic, and cultural factors and the personalities of participants. The first design consideration is to uncover and tailor to the diversity of participants. For example, Frei et al. (2019) grouped OA based on their physical capability levels and allocated walking trails with appropriate intensity levels to each group (38). O’Brien et al. (2021) identified a few factors influencing whether tracking devices can promote physical activities for OA: self-awareness, previous engagement with self-tracking, and whether one is more intrinsically or extrinsically motivated (45). It has also been found that some OA prefer competition (45), while others prefer collaboration (42) when engaging in physical activities.
The second design consideration is to recognise any ageism views held by the stakeholders and researchers themselves. Sometimes OA can perceive these ageism views, while other times, they also have ageism views about themselves. The creation of radio stations, according to some OA, is an act that challenges ageism because it is an effective medium to democratise the voices of OA (49). Meanwhile, they advocated for an intergenerational network and considered intergenerational content choices to connect OA with society and let society know that OA is not a homogenous group. Lenstra (2017) observed that ageism views appear in policies and institutional structures and practices, which are usually created by some stakeholders (36). They found that, among OA, some exert their agency to work around these policies and practices or shape new ones, while others were surprised when they felt that society wanted them at their age. It is important for researchers to be aware of ageist thoughts when stakeholders are not.
In regard to involving technology in the intervention, it is key to know what technologies OA are familiar with and, if needed, include non-technical options as part of the intervention. For instance, Yang & John (2020) only deployed a TV display as the technological element of the Team Bingo game given what technologies OA are familiar with (39). When selecting self-threading machines for the makerspace, some OA stressed that they come to the makerspace to make things rather than to learn technologies; in response, the committee decided to purchase two machines with one set at the default setting so that OA can use it without interacting with the “technical bits” (44). Reuter & Liddle (2020) identified that radio content creation engages both non-digital (e.g., presenting) and digital (e.g., editing) skills, with broadcasts as a bridge to the digital divide (49). OA can start with what they are good at and work towards the same goal. The non-digital and human interactions during content creation also provide a safe and accessible environment for OA to learn and practice digital skills. They observed that all members began to engage digitally to stay involved with content creation during the COVID-19 pandemic.
Dissemination
This category aims to support researchers in communicating findings and knowledge gained to key stakeholders. Only Frei et al. (2019) reported the dissemination approach, yet we extracted three insightful design considerations from this study (38). The first design consideration is to provide periodical findings to stakeholders during intervention development. This is different from and complements the design consideration of “regularly collecting feedback and ideas from key stakeholders”. Together, these two design considerations foster two-way communication between the research team and the community. In addition to presenting the intervention development progress, the researchers can also reflect on the procedural knowledge with the community, for example, how the research team and the community can collaborate better.
The second design consideration is to create a practical guide to help communities implement the intervention. After evaluating the feasibility and acceptability of the intervention via interviews, Frei et al. (2019) developed a manual for intervention implementation, which can be disseminated to communities that have similar goals of promoting physical activity. This action scales the social impact of their research findings.
The third design consideration is to train the responsible person(s) for intervention handover in sufficient time, as knowledge can be both explicit and implicit. Implicit knowledge is hard to express in words but can be transferred by training, such as observing others performing a task, and early involvement helps with knowledge transfer (50). Frei et al. (2019) started the transfer five months before the end of the research, and they trained the responsible persons from organisation skills, technical know-how, and recruitment strategies to funding application techniques.