22. Integrity, Power & Justice in Research

Research as a political act

Research distributes benefits and costs. It allocates attention. It legitimates some framings and discredits others. These are political acts even when the researcher claims neutrality. The claim of neutrality is itself a political stance — it leaves dominant framings unexamined while presenting itself as outside politics.

The transdisciplinary discipline is to name the politics of the research, not to escape them. A study that names what it serves, whose interests it advances, and what it leaves unchallenged is more honest than one that doesn't.

This does not mean every study is advocacy. Some research is descriptive, some explanatory, some translational. The point is that every research design embeds choices — about question, method, audience, output — that have political consequences.

Beneficiary analysis

A beneficiary analysis asks three questions of any project:

1. Who benefits, and how much?
2. Who bears the costs, and how much?
3. What rebalancing would make the distribution fair?

The third question is the action question. Analysis without redistribution is description.

Common findings of beneficiary analyses:

• Benefits concentrated in academic stakeholders — publications, grants, tenure, career
• Costs distributed to participants and community partners — time, emotional labor, exposure
• Asymmetric risk — researchers face minimal personal cost; participants face exposure of stigmatized conditions
• Non-monetary costs — community partners spending political capital on the project

Naming the asymmetry doesn't automatically fix it. It does open the question of rebalancing: paid roles, co-authorship, community-controlled outputs, capacity-building, post-grant continuation.

Data justice

Data justice is a framework that extends ethics into questions of who has rights over data:

• Sovereignty — communities, especially those historically subject to extractive research, hold collective rights over data about them
• Reciprocity — data flows to the studied community in usable form, not only to researchers
• Refusal — communities may refuse particular analyses, even after consenting to data collection
• Accessibility — data should be usable by the community, not only by researchers with specialized tools

Data privacy protects individuals from disclosure. Data sovereignty grants collective rights. Both are required; they aren't the same thing.

A practical pattern: a Memorandum of Understanding that names data ownership, retention policies, secondary-use rules, and the community's right to refuse particular analyses. The MOU is part of the data infrastructure, not a courtesy.

Whose framing wins?

Research framings are not neutral. The choice of how to frame "the problem" advantages certain interventions and disadvantages others. Examples:

• Framing diabetes as a behavioral problem privileges individual education
• Framing diabetes as a structural problem privileges policy and environmental intervention
• Framing homelessness as a mental-health problem privileges treatment
• Framing homelessness as a housing problem privileges housing-first policy

Each framing is partial; each generates different research questions and policy implications. The political choice is which framing structures the project. A transdisciplinary frame considers multiple framings deliberately rather than defaulting to the disciplinary norm.

Power-sharing in research relationships

Power dynamics show up in research relationships:

• Between researcher and participant — the researcher has institutional resources, professional standing, and exit options the participant doesn't
• Between academic and community partner — the academic team typically holds the grant, the methodological authority, and the publication infrastructure
• Between funder and researched community — the funder's priorities often shape what gets studied
• Between disciplines — methodological hegemony favors quantitative approaches in many funders' rubrics

Power-sharing requires structural changes:

• Community partners as named PIs or co-PIs
• Budget lines for community partner organizations
• Decision authority shared on specified questions
• Conflict-resolution processes that don't default to academic hierarchy

Data sovereignty in indigenous contexts

For indigenous communities, data sovereignty has particular weight. CARE Principles for Indigenous Data Governance articulate four pillars:

• Collective benefit — data ecosystems should enable Indigenous Peoples to derive benefit from their data
• Authority to control — Indigenous Peoples' rights and interests in Indigenous data must be recognized
• Responsibility — those working with Indigenous data have a responsibility to share how those data are used
• Ethics — Indigenous Peoples' rights and wellbeing should be the primary concern

CARE complements rather than replaces FAIR (Findable, Accessible, Interoperable, Reusable) principles. Together they articulate that data should be both technically accessible and ethically governed.

The audit framework

A useful audit framework for the politics of a project:

• Question: who shaped the research question? Whose problem does it address?
• Methods: whose methodological tradition is privileged? Whose is treated as auxiliary?
• Sample: who is included? Who is missing? What does the inclusion/exclusion say about whose problem this is?
• Outputs: where do the outputs go? Who reads them? Who can use them?
• Authorship: who is credited? Whose contribution is acknowledged vs. omitted?
• Continuation: what happens to the relationship after the grant ends?

A study can fail any of these without failing IRB. Passing all of them takes intentional design.

A worked vignette

A team studying immigrant access to mental health services in a U.S. county runs a beneficiary analysis early in the project:

• Benefits: publications for the academic team, intervention prototype for the clinic system, public-health data for the county
• Costs: participants' time and emotional labor in sharing migration histories; potential exposure for undocumented participants; community partners' time in advisory work
• Rebalancing moves: paid community navigators rather than unpaid advisors; community-controlled data deletion protocols at study end; co-authorship for community partners; commitment to county-funded continuation of intervention if effective

The audit informed the proposal, not just the discussion section. The funder included the rebalancing line items in the budget.

When the political stakes are high

Some research happens in contexts of acute political contention — voting rights, abortion access, immigration enforcement, climate-related migration. The political stakes are part of the methodological choices.

Conditions to acknowledge explicitly:

• The findings will be used by actors with political agendas
• The research relationship sits inside a political context the researcher doesn't control
• The framing chosen advantages particular policy positions
• The dissemination strategy has political consequences

A research team that pretends political neutrality in these contexts produces work that gets used politically anyway, often without the team's input. Acknowledging the stakes and engaging deliberately is more honest.

Closing

Research is political; neutrality is a stance. Beneficiary analysis asks who benefits, who bears cost, and what rebalancing is required. Data justice extends ethics into rights over data, including sovereignty and refusal. Power-sharing requires structural commitments. Framing choices have political consequences and should be made deliberately.

Next: authorship, publication, and knowledge sharing — ethics of credit in team science and open-science practices.