Source-linked AI summary
A Systematic Review and Thematic Analysis of Community-Collaborative Approaches to Computing Research
Ned Cooper, Tiffanie Horne, Gillian Hayes, Courtney Heldreth, Michal Lahav, Jess Scon Holbrook, Lauren Wilcox
TL;DR
Computing lacks a cohesive understanding of community-collaborative research, despite its growing use in technology research and design. The authors systematically review and thematically analyze computing research to identify project dimensions and tensions, many concerning researchers’ power and position relative to community partners.
Problem
HCI and computing lack a cohesive understanding of community-collaborative research, while scholarship on it remains siloed from computing.
Source: §1 · p. 1Method
The authors conduct a systematic literature review and reflexive thematic analysis of computing research on community-collaborative approaches.
Source: §7 · p. 14Results
The review identifies seven dimensions of community-collaborative projects and tensions related to researchers’ positions and the computing research environment relative to community partners.
Source: §7 · p. 14Takeaways & Limitations
The authors call for structural changes, methodological proposals, and ongoing reflexive practice to support community-centered research.
Source: §6 · p. 11The review excludes gray literature, including community-produced work published outside peer-reviewed journals and conference proceedings.
Source: §3.1.1 · p. 4
Abstract
from arXiv · showhide
HCI researchers have been gradually shifting attention from individual users to communities when engaging in research, design, and system development. However, our field has yet to establish a cohesive, systematic understanding of the challenges, benefits, and commitments of community-collaborative approaches to research. We conducted a systematic review and thematic analysis of 47 computing research papers discussing participatory research with communities for the development of technological artifacts and systems, published over the last two decades. From this review, we identified seven themes associated with the evolution of a project: from establishing community partnerships to sustaining results. Our findings suggest that several tensions characterize these projects, many of which relate to the power and position of researchers, and the computing research environment, relative to community partners. We discuss the implications of our findings and offer methodological proposals to guide HCI, and computing research more broadly, towards practices that center communities.
1 INTRODUCTION
Although participatory approaches have a history in computing, scholarship on community collaboration remains siloed from HCI and computing, leaving the field without a cohesive understanding of these approaches. This paper reviews prior work, identifies themes and tensions across projects, and offers reflections and proposals for research that centers communities.
1 INTRODUCTION
- Scholarship on community collaboration remains siloed from HCI and computing, risking extractive research without mutual benefit or attention to inequities.The authors identify this as a reason for grounding computing research in the relevant body of work.Source: §1 · p. 1
- The authors use community-collaborative approaches (CCA) for research involving community members as research-team participants throughout technological research and development.The umbrella term encompasses approaches such as Action Research and Community-Based Participatory Research.Source: §1 · p. 1
- The review asks how computing researchers involve communities as co-researchers, what collaboration elements are reported, and how publications align with community-centered approaches.It also asks what critiques and methodological innovations follow from divergences from those approaches.Source: §1 · p. 2
- The paper combines a meta-review with reflexive thematic analysis to identify seven themes and tensions across project stages.It also reflects on findings and proposes methodological approaches for HCI and computing research.Source: §1 · p. 2
- The findings call for attention to researchers’ power and position and to the computing research environment relative to community partners.The authors frame the paper as a starting point for researchers engaging community stakeholders and propose structural and methodological changes to support meaningful, mutually beneficial participation.Source: §1 · p. 2
2 BACKGROUND
The background traces participatory research in HCI from Action Research and Participatory Design to community-based approaches, while emphasizing that participation raises questions of power, benefit, and inclusion. It presents community collaboration as a broader commitment to work with communities throughout research, not simply to solicit individual feedback.
2.1 HCI and Action Research
- Action Research emphasizes collaboration with community stakeholders to co-construct knowledge for real-world problems, while Participatory Design developed from workplace efforts to involve workers in system design.HCI adopted Participatory Design as an applied design-research approach, and projects have since moved beyond workplaces and studios into varied contexts and community spaces.Source: §2.1 · pp. 2–3
- Many Participatory Design projects involve individuals’ creative input or feedback, sometimes in workshops away from community sites and with participants identified by researchers.The passage contrasts this pattern with projects led by historically marginalized groups and approaches that work in community spaces.Source: §2.1 · p. 3
2.2 The Emergence of Community-Based Approaches in HCI
- CCA encompasses Action Research, CBPR, and related efforts involving community stakeholders as co-researchers throughout research on community concerns and technological or sociotechnical solutions.The approaches aim to cultivate embedded partnerships, foreground collective community voice, and engage community and researcher epistemologies.Source: §2.2 · p. 3
- CCA can shift influence from an intervention acting on a community to a community influencing the intervention, drawing on members’ lived experience in technology development.The text says this can increase product relevance, surface potential negative impacts, and mitigate harmful impacts on marginalized communities.Source: §2.2 · p. 3
2.3 Debates About Participation
- Participation is not an unqualified good: requiring visible inclusion can tokenize community members or pressure people to disclose stigmatized identities.The authors advocate participation to the degree useful and productive for community members, alongside transparency that supports informed choices.Source: §2.3 · p. 4
3 METHOD
The section details a systematic review for data collection and reflexive thematic analysis, then situates the findings in the researchers’ positionality and the review’s limitations.
3 METHOD
- The review combined systematic data collection with reflexive thematic analysis, its primary contribution.The authors also contextualize their analysis through positionality and discuss limitations of the approach.Source: §3 · p. 4
3.1 Data Collection
- Data collection followed PRISMA-informed steps: database searches, abstract and full-text screening, eligibility checks, and corpus extraction.The search covered five databases, and two authors independently reviewed full texts before eligibility decisions.Source: §3.1 · p. 4
3.1.1 Information Sources and Search Strategy.
- Searches targeted community-based technological development in five databases, using peer-reviewed English-language articles from 2000 through July 2021.The review excluded gray literature and notes limited coverage of non-English venues.Source: §3.1.1 · pp. 4–5
3.1.2 Screening and Eligibility.
- Eligibility required community members to participate in developing a technological artifact or system; screening left 47 papers for analysis.The authors excluded work that studied communities or existing technologies without developing an artifact or system with the community.Source: §3.1.2 · p. 5
3.1.3 Data Extraction.
- For each paper, the authors recorded publication characteristics, participating communities and actors, research approach, technology, outcomes, and project duration.The extraction included quantitative and qualitative data.Source: §3.1.3 · p. 5
3.2 Data Analysis
- Two authors inductively developed and revised codes through repeated independent reading, discussion, and reflection, then collaboratively generated themes.They used flexible coding rather than predefined themes or inter-rater reliability testing, and discussed explanatory memos with other authors.Source: §3.2 · pp. 5–6
3.3 Positionality of the Research
- The authors acknowledge that their Western institutional positions and status as computing researchers, rather than community participants, shape their interpretations.They also note that the reviewed corpus represents CCA predominantly from a Western perspective.Source: §3.3 · p. 6
3.4 Limitations
- Synthesizing across studies limited the depth available for any one paper, while descriptive statistics reflect what publications reported rather than all project outcomes.The authors suggest narrative reviews may better identify best practice in individual projects, while their synthesis can reveal cross-project themes and trends.Source: §3.4 · p. 7
4 FINDINGS: DESCRIPTIVE STATISTICS
The review describes the corpus by publication year, venue, location, community, research approach, technology, and project duration. Most papers appeared in the final five years surveyed, and projects varied in setting and length.
Publication year and venue
- 29 papers (62%) were published in 2017–2021, the final five years of the surveyed span; the corpus covered 14 of 22 years.Articles appeared most often in CHI (8), CSCW (6), and DIS (5).Source: §4 · p. 7
Geographical location
- About half of the articles described projects in the U.S. or Canada (24, 51%); 13 (28%) were in the Global South, six (13%) in Europe, and three (6%) in Australia.The Global South projects were situated in Africa and Asia.Source: §4 · p. 7
Community
- Table 1 counts articles directly involving each community category; some articles belong to more than one category, and labels reflect terms reported in the reviewed papers.The review also describes the corpus across community participants and project types.Source: §4 · p. 7
Stated research approach
- Twenty-seven articles (57%) used participatory-research labels while describing community-based project elements; 20 (43%) used community-based research labels.The paper grouped 13 reported approaches into two categories; its listed terms include participatory and community-based variants.Source: §4 · p. 7
Types of technology developed
- The corpus included varied technologies, from mobile and computer apps and websites to robots, drones, and systems for knowledge management and air-quality monitoring.Other examples included a system for co-operative canoe paddling.Source: §4 · pp. 7–8
Time frame
- Projects lasted one month to one year in 20 cases (43%) and one or more years in 17 (36%); nine (19%) lasted less than a month.One article did not report project length.Source: §4 · p. 8
5 FINDINGS: DIMENSIONS AND TENSIONS OF COMMUNITY-COLLABORATIVE PROJECTS
The findings organize community-collaborative projects into seven dimensions, understood as elements contributing to holistic, participatory engagement. Across these dimensions, tensions can arise from researchers’ and the computing environment’s power and position relative to communities.
Dimensions and tensions
- The seven dimensions are elements or factors contributing to holistic, participatory engagement by researchers and communities in a project.The findings structure these dimensions around a project's evolution, from partnership formation to sustaining results.Source: §5 · p. 8
- Tensions can arise within project dimensions from researchers’ and the computing research environment’s power and position relative to participants and community members.Source: §5 · p. 8
- The findings follow project evolution from establishing a partnership through participatory decision-making and collaborative activities to sustaining results.Source: §5 · p. 8
5.1 Establishing a Partnership
Establishing a partnership involves identifying community partners, building relationships, and deciding when and where a project begins. The reviewed projects commonly worked through existing community organizations and situated research in community settings.
5.1.1 Defining a Community Partner.
- In 31 projects (66%), researchers connected with participants through an existing community-based organization or representative group, while CCA encourages relying on community self-definition.Some researchers were approached directly by communities, while others recruited through advertisements and community gathering places.Source: §5.1.1 · p. 8
- Most projects involved users, intended users, or artifact maintainers as participants (44, 94%); 23 (49%) also included other community members, and nine (19%) included non-users.Source: §5.1.1 · p. 8
- Some projects included community representatives on the research team or as co-authors; one example followed a seven-year project in Namibia from a researcher native to that community.Other projects included community members as article co-authors.Source: §5.1.1 · p. 8
5.1.2 Building Relationships.
- Thirty-four projects (72%) reported relationship-building before research activities, and researchers without prior ties described sustained efforts to build trust with participants and wider community members.Examples included an 18-month effort to establish a community advisory board and immersion in community settings.Source: §5.1.2 · pp. 8–9
- Communities initiated 14 projects (30%); researchers initiated 33 (70%), commonly with some collaboration, and nearly all projects were situated in community settings for at least one stage (46, 98%).Among researcher-initiated projects, 21 (45%) involved communities during planning and 12 (26%) during activities; 26 projects were situated in communities throughout and 18 for multiple stages.Source: §5.1.2 · p. 9
5.2 Enabling Participatory Models for Decision-Making and Collaborative Research
The findings describe community knowledge informing design, uneven sharing of control, and iterative, adaptive collaboration. These practices can require substantial community commitment while researchers also face pressure to produce academic and technological outputs.
5.2.1 Integrating Community Knowledge and Practices.
- Twenty-six projects (55%) reported integrating community knowledge or practices into design, in some cases allowing local customs and values to shape the process and decision-making.Examples included using community networks and holding activities at customary places or times.Source: §5.2.1 · p. 9
- Although 44 projects (94%) described collaborative ideation, communities rarely helped conceptualize the problem; participants also joined prototype development (26, 55%) and critique (23, 49%).Usability critiques of researcher-conceived problems do not by themselves establish a solution’s relevance or value to a community.Source: §5.2.1 · p. 9
5.2.2 Sharing Control.
- Researchers and communities shared control of planning and managing research activities in 33 articles (70%), and of artifact development in 27 (57%); only four (8%) discussed shared control of data analysis.Source: §5.2.2 · pp. 9–10
- Twenty-seven articles (57%) described efforts to address power differences by elevating community voices, including adapting methods to support participation.One project created a multisensory toolkit so visually impaired children could participate in prototyping alongside classmates.Source: §5.2.2 · p. 10
5.2.3 Cycles of (Re)Production and Adaptive Collaboration.
- Most projects used multiple phases with community partners (31, 87%), and 31 (66%) reported adapting methodology as collaboration developed.Adaptations included changing plans and goals, flexible workshop formats, storytelling, and fitting activities to residents’ daily rhythms.Source: §5.2.3 · p. 10
- Iteration may transfer knowledge between partners but also increases community members’ commitment; required academic and technology outputs can limit researchers’ openness to community-led methods.The authors call attention to the broader cycles and affective demands of outsiders’ engagements with marginalized communities.Source: §5.2.3 · p. 10
5.3 Sustaining Results of a Project
Projects often describe benefits for both researchers and communities, but those benefits may arrive at different times and are rarely measured for communities. Papers also report efforts to share results and sustain artifacts, while post-project relationships receive less attention.
5.3.1 Ensuring Mutual Benefits.
- Outputs were identified for both researchers and communities in 29 projects (62%), but community benefits were often incremental, delayed, or unevenly experienced.Few papers described approaches to measuring community benefits, which may reflect difficulty bridging researcher and community measures of success.Source: §5.3.1 · pp. 10–11
- Across the projects, capacity building was identified as a benefit in 31 (67%), particularly community advocacy capacity, while researchers’ understanding of partner communities and context was reported in 39 (83%).Source: §5.3.1 · p. 10
5.3.2 Sharing Results and Maintaining Relationships.
- Researchers partnered with participants to share results with community members or wider stakeholders in 18 projects (38%).Publications alone could not confirm whether reported absences of sharing steps meant those steps were not taken in practice.Source: §5.3.2 · p. 11
- Sustainability of an artifact or system was considered in 31 projects (66%), including community training and plans for ongoing use, ownership, or funding.Some papers described training community members to operate and manage systems, while others raised the need for sustainable structures and continued funding.Source: §5.3.2 · p. 11
- Few papers reported how researchers would maintain partner-community relationships after their substantive engagement ended.Community organizations may change or dissolve, requiring ongoing relationship maintenance.Source: §5.3.2 · p. 11
6 DISCUSSION
The discussion identifies a gap between community-collaborative principles and how computing research publications report projects, alongside structural constraints on researchers. It proposes methodological changes and reflexive practices to better center community participation.
6 DISCUSSION
- The discussion frames its response as structural change, four methodological proposals, and ongoing researcher reflexivity.It proposes developing a community of practice around community-collaborative approaches.Source: §6 · p. 11
6.1 Structural Changes
- Most reviewed publications described community involvement piecemeal: communities often entered after problem definition, and researchers usually led data analysis.Few examples described shared analytic control; the authors argue that research and publication structures should support community involvement in defining questions and analyzing data.Source: §6.1 · pp. 11–12
6.2 Methodological Proposals
- The authors identify methodological innovation as a parallel to the broader structural research agenda.The proposals address community definition, contestability, community knowledge and practices, and sustaining results.Source: §6.2 · p. 12
6.2.1 Defining Communities.
- Defining a community by researcher-selected common traits can overlook collective identity and internal differences, and may disempower the group.For marginalized communities, defining community on their own terms may reassert power; online networks also pose challenges for identifying and connecting communities.Source: §6.2.1 · p. 12
6.2.2 Facilitating Contestability of HCI Research and Design Prac-
- Consensus-focused research and publication can obscure disagreements that contribute to knowledge and change in community-collaborative projects.The authors propose that HCI researchers facilitate forums where communities can contest technology designs and consider alternatives.Source: §6.2.2 · pp. 12–13
6.2.3 Building from Community Knowledge and Practices.
- Researchers often adapted methods to community settings, but few reported substantially reorienting research and design practices around community knowledge.The authors call for reporting on methodological pivots and using community engagement to reflect on and improve HCI methods.Source: §6.2.3 · p. 13
6.2.4 Sustaining Results.
- The review found limited guidance on defining and measuring benefits for communities and individual participants, and few evaluations over time.The authors propose measures that include well-being, relationship stability, and community outcomes, plus incentives for ongoing evaluation and reporting.Source: §6.2.4 · p. 13
6.3 Reflexive Practice
- Reflexive practice includes considering historical trust in research institutions and the systems connected to a project, as well as documenting tensions and relationship work.The authors suggest that documentation protocols can increase transparency and support critical reflection and a community of practice.Source: §6.3 · pp. 13–14
7 CONCLUSION
The review identifies seven dimensions of community-collaborative projects and tensions shaped by researchers’ positions and computing research environments. It calls for addressing systemic issues alongside methodological innovation to support responsible, mutually beneficial projects.
7 CONCLUSION
- The review identifies tensions around researchers’ positions and computing research environments relative to community partners, while noting that institutional cultures and publication conventions constrain practice and reporting.The authors do not intend the review to resolve these tensions and note that incentives and power imbalances will remain.Source: §7 · p. 14
- The authors call for studying and addressing systemic issues that affect computing research’s ability to achieve community-collaborative goals.They hope this agenda, alongside methodological innovations, will help responsible and mutually beneficial projects become central to HCI and computing research.Source: §7 · p. 14