03 · What You Need to Know
Data Collection Is Part of the Research the Committee Reviews
Prospective Review Means Review Before the Research Activity
Research ethics review is prospective for a reason. The committee evaluates the proposed study before participants are exposed to its procedures or their information is used in ways requiring prior review.
Under U.S. HHS requirements, investigators should ensure that activities meeting the definition of nonexempt human-subject research undergo IRB review and approval before those activities are initiated. OHRP's training materials likewise state that initial IRB review occurs before human-subject research activities are allowed to begin.
The 2022 Philippine National Ethical Guidelines similarly frame ethical review and clearance as occurring before implementation of covered research involving human participants.
Data Collection Begins Earlier Than Researchers Sometimes Think
Researchers may imagine “data collection” as the moment completed questionnaires begin arriving or laboratory measurements are entered into a spreadsheet. In practice, research information can be obtained through many activities.
| Activity |
Can It Involve Research Data? |
| Administering a questionnaire |
Yes. Participants are providing information for the research. |
| Conducting an interview or focus group |
Yes. Statements, recordings, transcripts, and field notes can be research data. |
| Performing measurements or assessments |
Yes. Scores, observations, images, physiological measures, and other results can be research data. |
| Observing participants |
Yes. Behavioral observations and research field notes can constitute data. |
| Accessing identifiable private records |
Potentially. Secondary research can involve identifiable private information even without direct participant contact. |
| Collecting eligibility information |
Potentially. Screening can involve identifiable and sensitive information from prospective participants. |
The relevant question is therefore not simply whether your “main outcome” has been measured. Ask whether a covered research activity involving people or their information has begun.
Waiting to Analyze the Data Does Not Make Early Collection Acceptable
A common proposed workaround is: collect the data now, keep the file closed, and analyze it only after approval.
That misses what prospective review protects. The ethics committee may need to evaluate recruitment, consent, questionnaires, interviews, risks, privacy protections, data security, interventions, participant selection, or other procedures before they occur.
If those activities have already happened, postponing statistical analysis does not recreate the opportunity for prospective review.
Data collection
Obtaining information, measurements, observations, recordings, specimens, or other research material from or about participants.
Data analysis
Examining or processing collected information to answer the research question.
Ethics approval can be required before the first of these, not merely before the second.
Recruitment Can Start Even Before Data Collection
Researchers should also avoid treating data collection as the first ethically relevant event. Recruitment can precede it and may itself require prospective authorization.
Advertising a study, directly inviting participants, collecting expressions of interest, and screening potential participants can all form part of the human-participant research process.
This is why participant recruitment while ethics approval is pending should be considered separately rather than assuming that everything before the questionnaire is merely administrative.
Screening Data Can Still Be Data
Suppose you do not administer the main questionnaire but ask potential participants to complete a screening form containing age, diagnosis, medication use, employment information, academic status, or another eligibility characteristic.
You have obtained information about identifiable or potentially identifiable people because of their prospective research participation. Depending on the study and applicable framework, those screening procedures may themselves need to be included in the approved protocol.
Do not assume that information from people who ultimately prove ineligible somehow never counted.
Existing Records Can Also Be Collected for Research
No direct participant interaction is required for premature data collection to occur. A researcher might begin extracting information from medical records, student files, administrative databases, or other identifiable private records while the ethics application remains pending.
If that activity constitutes covered secondary research, waiting to contact participants is irrelevant. The research use of the records has already begun.
The applicable requirements for research using existing data should therefore be settled before accessing or extracting information for the study when prior review is required.
Downloading Public Data Can Be Different
Not every activity called data collection involves covered human-participant research. Researchers may download genuinely public aggregate statistics, use synthetic data, or work with information falling outside a particular human-research framework.
In such cases, the ethics requirements can differ substantially. The key is to establish the status of the activity rather than assuming that every spreadsheet requires IRB approval or, at the other extreme, that every dataset is safe to collect before review.
If you are relying on public availability, determine whether the publicly available data actually alter the ethics-review requirement.
Can You Pilot the Instrument While Approval Is Pending?
That depends on what “pilot” means.
Testing survey logic with synthetic responses, asking members of the research team to click through a form, checking skip patterns, rehearsing an interview procedure among investigators, or testing recording equipment without collecting research information about people can be preparatory activities.
Recruiting real participants to complete the questionnaire, collecting their responses, testing an intervention on them, or systematically obtaining information from them for the study is different. A pilot involving human participants can itself require an ethics determination.
Calling the first few cases “test data” does not make those participants less real.
Conditional or Pending Approval Is Not Automatically Permission to Collect Data
An ethics committee may request revisions or approve a protocol subject to specified conditions. Researchers should follow the formal decision and institutional procedure for determining when approval becomes effective.
OHRP distinguishes approval with conditions from situations in which an IRB cannot yet make all determinations required for approval. In either case, researchers should not infer permission to begin merely because the outstanding changes appear minor.
If the approval letter does not authorize data collection, do not manufacture authorization from optimism.
Why Waiting Matters Even for Minimal-Risk Research
Minimal-risk research can still involve issues that the ethics process may change. Reviewers may request removal of unnecessary identifiers, revisions to sensitive questions, changes to recruitment, stronger confidentiality protections, a different consent process, or additional safeguards for participants in dependent relationships.
If data have already been collected using the original version, those protections cannot be applied prospectively to the people who already participated.
Preparing for Data Collection Is Different From Collecting Data
There is usually plenty researchers can do while ethics review is pending, subject to institutional rules.
Preparation
Drafting instruments, programming forms, training research staff, testing systems with synthetic information, preparing codebooks, developing analysis scripts, and organizing logistics.
Human-participant research activity
Recruiting, screening, consenting, intervening with, observing, or obtaining covered information or materials from or about actual participants for the study.
The efficient approach is to make the project ready to launch, not to launch part of the project before authorization.
Funding and Deadlines Do Not Change the Ethics Requirement
A funding period may have started. A thesis deadline may be close. Field access may exist for only two weeks. Participants may be unusually available today.
Those are real project-management problems, but they do not transform pending approval into effective approval.
If timing is critical, contact the ethics office about its procedures rather than beginning data collection independently.
Do Not Assume Approval Will Cover Data Collected Earlier
An approval issued on a later date ordinarily authorizes the research according to its terms from the effective point established by the ethics body. Researchers should not assume that it automatically reaches backward and validates research activities conducted before approval.
The separate question of whether ethics approval can be obtained retrospectively is considerably more difficult and should not be used as a planned strategy.
Watch Out
Do not collect participant responses, interviews, measurements, identifiable records, specimens, or other covered research data “at your own risk” while waiting for approval. The risk is not merely administrative: the prospective participant protections being reviewed may no longer be possible to implement for those already included.
