03 · What You Need to Know
Sharing Data Collection Decisions With Community Research Partners
What Does Community Involvement in Data Collection Decisions Mean?
Community involvement in data collection decisions refers to participation by community representatives or research partners in determining the information a study seeks, how that information is defined, and whether proposed collection procedures are appropriate.
This involvement may occur through advisory consultation, instrument co-development, shared governance, or formally delegated responsibilities.
For example, community partners may help identify relevant experiences for an interview schedule or determine whether a questionnaire adequately represents local conditions.
Their involvement should be distinguished from simply collecting data on behalf of researchers. A community member who administers a completed questionnaire may have an operational role without having influenced what the questionnaire measures.
Meaningful participation requires clarity about which decisions partners can influence and how their recommendations affect the final protocol.
Why Might Researchers Overlook Important Data?
Researchers often develop instruments from established theoretical frameworks, previous studies, administrative categories, or available datasets.
These sources can provide valuable methodological foundations, but they may not capture every relevant feature of the population or setting.
For example, researchers studying digital access may measure whether students own a laptop and have an internet connection.
Community partners may explain that device sharing, electricity interruptions, mobile data costs, and accessibility software also influence participation.
Such contributions may reveal weaknesses in the operational definition of digital access.
However, community experience should be evaluated alongside other evidence. A single representative's observations should not automatically be treated as characteristic of every member of the population.
Which Data Collection Decisions Can Community Partners Influence?
| Decision |
Potential community contribution |
Methodological consideration |
| Construct definition |
Clarify what a concept means in the local setting. |
Conceptual validity and alignment with the research question |
| Variable selection |
Identify important characteristics or experiences. |
Analytical necessity and construct coverage |
| Questionnaire content |
Review relevance, wording, and missing domains. |
Measurement validity and comparability |
| Interview questions |
Suggest appropriate topics and terminology. |
Qualitative research purpose and ethical sensitivity |
| Sensitive information |
Identify possible privacy or social risks. |
Necessity, consent, confidentiality, and data minimization |
| Data collection mode |
Recommend accessible and acceptable procedures. |
Mode effects, feasibility, and participant selection |
| Data governance |
Contribute to appropriate stewardship and use arrangements. |
Legal authority, consent, and recognized governance rights |
| Secondary data use |
Identify concerns about purposes beyond the original study. |
Authorization, compatibility, and applicable restrictions |
The appropriate degree of influence depends on the partnership agreement and the scientific and institutional responsibilities attached to each decision.
Can Community Partners Help Define Research Variables?
Yes. Community partners may identify differences between researchers' conceptual definitions and the experiences those definitions are intended to represent.
For example, a study of household food security might initially emphasize food availability. Community partners may identify affordability, accessibility, or culturally appropriate food as relevant dimensions.
Researchers should examine whether these dimensions belong to the intended construct or represent related but distinct variables.
Community involvement can improve conceptual clarity, but it does not automatically establish that every proposed dimension should be combined into one score.
Construct definition should be supported by theory, relevant empirical evidence, and the intended interpretation of measurements.
Can Community Partners Request Additional Variables?
Yes. They may identify contextual characteristics, experiences, or potential explanatory factors missing from the original data collection plan.
For example, community partners studying access to public services may recommend collecting information about travel time, informal fees, or previous experiences with service providers.
Researchers should evaluate whether the proposed variables address the research question, support planned analyses, or provide important contextual information.
Additional variables may also increase participant burden, missingness, confidentiality risks, and analytical complexity.
The research team should therefore establish a rationale for collecting each consequential variable rather than expand the instrument indefinitely.
Can Community Partners Request Removal of Sensitive Questions?
They can raise concerns and, where authorized, participate in decisions about whether sensitive questions should be included.
Sensitivity may arise from the subject matter, social context, possibility of identification, or consequences of disclosure.
For example, questions about income, immigration circumstances, family relationships, or stigmatized experiences may create different risks across settings.
Researchers should evaluate whether the information is necessary and whether less intrusive measures can support the intended analysis.
However, removing a variable may also affect the study's ability to address confounding, describe relevant inequalities, or answer the research question.
The appropriate response is to examine scientific necessity and ethical consequences together.
What Is the Difference Between Data Relevance and Data Necessity?
Data relevance
Concerns whether information relates meaningfully to the phenomenon, population, or research purpose.
Data necessity
Concerns whether collecting particular information is justified for the specified research purpose and cannot reasonably be avoided or replaced under applicable requirements.
Information may be relevant without being necessary to collect from every participant.
For example, detailed household financial information may help explain educational access, but the study may be able to address its question using less intrusive indicators.
Conversely, an essential confounder or outcome measurement may require information that participants find sensitive.
Researchers should justify collection according to the study's needs and evaluate proportionate safeguards.
How Does Data Minimization Apply to Community Research?
Data minimization generally concerns limiting personal data collection and processing to information appropriate and necessary for specified purposes, subject to the applicable legal framework.
Its precise legal requirements differ across jurisdictions.
Community partners may help identify questions whose sensitivity or burden is disproportionate to their research value.
However, minimizing data does not mean removing every sensitive variable or collecting so little information that the research becomes scientifically uninterpretable.
Researchers should establish a defensible balance between analytical requirements and participant protection.
Where appropriate, they may consider less granular categories, restricted access, separation of identifiers, or other safeguards.
Can Community Partners Help Develop Questionnaires?
Yes. Partners may review item wording, response categories, examples, and the relevance of proposed constructs.
They may identify unfamiliar terminology, assumptions about institutional practices, or experiences not represented in the instrument.
For example, a questionnaire measuring digital learning access might ask whether a student has a personal computer. Community partners may explain that shared-device arrangements are common and require separate measurement.
However, changing an established instrument can alter reliability, factor structure, content validity, and score comparability.
Researchers should evaluate adapted instruments using appropriate measurement procedures rather than assume that stakeholder approval establishes validity.
Can Community Partners Help Design Qualitative Interview Questions?
Yes. Community knowledge can be particularly useful when researchers seek to understand experiences, practices, or institutional relationships.
Partners may identify topics that conventional interview schedules overlook or recommend language that participants can understand.
They may also identify questions likely to create discomfort or misunderstanding.
However, interview schedules should remain aligned with the qualitative research question and methodological approach.
Researchers should distinguish necessary flexibility in qualitative interviewing from unplanned changes that undermine comparability or leave essential topics unexplored.
Where community partners conduct interviews, appropriate training, confidentiality arrangements, and reflexive consideration of their relationships with participants are important.
Can Community Partners Help Determine Data Collection Frequency?
Yes. Repeated data collection can impose substantial demands on participants.
Community partners may identify scheduling conflicts, cumulative burden, or circumstances affecting follow-up.
For example, a longitudinal educational study might propose weekly surveys during examination periods, when students already face substantial academic demands.
However, reducing measurement frequency may prevent researchers from observing changes at the temporal resolution required by the research question.
The team should evaluate the consequences for longitudinal inference, missing data, and measurement validity.
These decisions are closely related to stakeholder assessment of participant burden.
Can Community Partners Help Decide How Data Are Collected?
Yes. They may recommend online, in-person, telephone, paper-based, or other appropriate procedures based on accessibility and participant circumstances.
However, data collection modes may affect who participates and how responses are recorded.
For example, a sensitive questionnaire completed privately may produce different response patterns from one administered by an interviewer known to the participant.
Researchers should evaluate mode effects, privacy, standardization, and sampling implications before adopting different procedures.
Where multiple modes are used, their comparability should be assessed according to the intended measurement.
Should Community Partners Be Able to Veto Particular Data Collection Activities?
They may possess approval or blocking authority over specified activities through legitimate governance rights or partnership agreements.
For example, a community governing body may control access to particular community-held records or require approval for collecting information governed by an established protocol.
However, community authority should not automatically be interpreted as control over every individual's participation decisions.
The scope of any veto should be explicit and consistent with applicable law, consent, ethics, and institutional requirements.
The distinction is examined in community research partner veto power.
How Do Indigenous Data Governance Principles Affect Collection Decisions?
Research involving Indigenous peoples may require attention to collective rights, cultural knowledge, community governance, and data sovereignty.
The CARE Principles for Indigenous Data Governance emphasize collective benefit, authority to control, responsibility, and ethics.
In the Canadian First Nations context, the OCAP principles address ownership, control, access, and possession of information.
However, these frameworks should not be treated as universally interchangeable or automatically applicable to every Indigenous community.
Researchers should work with the relevant governing institutions and follow applicable legal requirements, community protocols, and negotiated agreements.
Community governance may affect which information can be collected, how it is stored, who may access it, and the purposes for which it may be used.
Can Community Partners Access Identifiable Research Data?
Potentially, when access is appropriately authorized and necessary for their agreed responsibilities.
However, partnership status alone does not confer unrestricted access to identifiable information.
Access should follow consent arrangements, confidentiality obligations, applicable data protection law, ethics approvals, and relevant community governance agreements.
Researchers should distinguish involvement in deciding what data are collected from permission to view or process the resulting individual records.
Where appropriate, partners may participate in analysis using de-identified or aggregated information rather than identifiable datasets.
Any proposed access arrangement should be established before data are shared.
What If Community Preferences Conflict With Essential Measurements?
Some variables may be necessary for the study's primary outcome, causal identification strategy, or other essential inference.
For example, a community partner may request removing questions about prior educational attainment from an observational study, while researchers consider that information important for addressing confounding.
The team should explain the variable's role and investigate whether a less intrusive but adequate alternative exists.
However, an alternative measure should not be accepted merely because it is easier to collect.
If essential information cannot be collected ethically or acceptably, the study may need to revise its research question, analytical strategy, or design.
Researchers should not claim that an analysis addresses confounding when the necessary information is unavailable.
Can Community Partners Help Decide Which Existing Data Sources Are Used?
Yes. Partners may identify relevant administrative records, community datasets, institutional reports, or other sources that researchers do not know exist.
They may also identify limitations in how those records were originally created.
For example, administrative data may record service use but omit unsuccessful attempts to access a service.
However, availability does not establish validity or authorization for research use.
Researchers should evaluate data provenance, definitions, completeness, measurement quality, and lawful access before incorporating a source.
Secondary use should remain consistent with applicable consent and governance requirements.
How Should Proposed Data Collection Changes Be Evaluated?
Each consequential proposal should be assessed according to its purpose and implications.
Researchers should consider whether the change improves construct coverage, reduces unnecessary burden, affects measurement comparability, or introduces new confidentiality risks.
Changes to primary outcomes, essential covariates, or core instruments may require revision of the analysis plan and additional methodological review.
Where uncertainty remains, pilot testing or cognitive interviewing may help determine whether the proposed measure works as intended.
However, a successful pilot does not automatically establish every required measurement property.
Decisions should be documented before implementation, particularly when the study involves several sites or longitudinal measurement.
Should Community Partners Help Interpret the Data They Requested?
They may provide valuable contextual knowledge about why particular variables matter and how observed patterns relate to community experiences.
However, interpretation should remain consistent with the data and research design.
For example, partners may explain why connectivity costs could influence digital learning participation, but the study must contain appropriate evidence before concluding that costs caused an observed difference.
Community contributions can help identify alternative explanations and limitations without determining findings by consensus.
The distinction between data collection decisions and community involvement in interpreting findings should remain clear.
How Should Community Contributions Be Documented?
Documentation should identify which partners participated, what changes they proposed, the rationale for those changes, and the final decisions.
It should also explain why recommendations were rejected or modified, particularly when scientific or ethical requirements prevented their adoption.
Where partners have formal decision authority, the relevant approvals and responsibilities should be recorded.
The GRIPP2 reporting checklists can support transparent reporting of patient and public involvement, although the appropriate scientific reporting framework depends on the study design.
Researchers should not claim that data collection was co-designed when partners merely reviewed completed instruments without a genuine opportunity to influence them.
Watch Out
Do not assume that community relevance automatically establishes measurement validity or that community approval authorizes unrestricted collection of sensitive information. Data collection must satisfy both scientific requirements and applicable participant protection and governance obligations.
