Can the Scope Change After the Study Has Started?

What Happens When the Study You Planned Is No Longer the Study You Can Conduct?

You begin data collection and discover that one population cannot be recruited. A data source you expected to use becomes unavailable. An instrument does not work as anticipated. A policy change alters the research context. Perhaps early fieldwork reveals that the original question was based on assumptions that do not hold.

Do you have to continue with the original scope simply because the study has already started?

Not necessarily. Research protocols do change after studies begin, and some changes may be scientifically, ethically, or practically necessary. In formal research settings, important changes are commonly handled as protocol amendments rather than treated as inherently prohibited. Reporting standards for trials likewise explicitly require disclosure of important methodological and outcome changes made after commencement and the reasons for them.

But changing scope after a study starts is not equivalent to refining a proposal before data collection. The later the change occurs, especially once researchers have seen data or emerging results, the greater the need to consider approvals, bias, comparability, analytical consequences, and transparent reporting.

Yes, but a Mid-Study Scope Change Must Be Treated as a Real Change

Scope Can Evolve, but the History of the Study Should Remain Visible

Not every change is a change in scope

Research projects change in many ways. Some modifications are administrative or procedural and leave the substantive inquiry essentially intact. Others alter what population, phenomenon, outcome, setting, timeframe, or question the study actually investigates.

Suppose an interview originally scheduled for 45 minutes is rescheduled to a different room. That is unlikely to change the substantive scope. By contrast, deciding halfway through recruitment to exclude an entire participant group can change the population represented by the study.

Similarly, correcting a typographical error in a questionnaire is different from replacing the primary outcome with another construct. Extending a recruitment deadline is different from expanding the population to a new institution whose context was not part of the original inquiry.

A useful question is: Does this modification change what the study is substantively trying to answer or the evidence used to answer it?

The exact classification depends on the study. A date change that is trivial in one project could be substantive in another if the phenomenon is highly time-sensitive.

Why might a legitimate scope change become necessary?

Research planning occurs before researchers know everything that will happen during implementation. Recruitment assumptions can fail. Access to records can change. Equipment can become unavailable. External events can alter the study environment. Early fieldwork can expose assumptions that need reconsideration.

Recent evidence from randomized trials illustrates that methodological changes after ethical approval are not unusual. A 2026 review of 100 NIHR-funded randomized trials identified 846 methodological changes, with reported reasons including feedback from recruiting sites and participants, logistical difficulties, incorrect planning assumptions, and changes needed to align with evolving clinical guidance.

The existence of such changes does not mean any modification is harmless. It shows why research governance includes mechanisms for amendments rather than assuming that every protocol will remain untouched from approval to completion.

The timing of the change matters

Consider three moments:

Before data collection. The researcher realizes that the original scope is unrealistic and revises it before participants are recruited or evidence is collected.

During data collection but before examining relevant outcomes. A recruitment or data-access problem forces reconsideration while the researcher remains unaware of how the change will affect the eventual result.

After examining the data or emerging results. The researcher knows that changing an outcome, population, subgroup, exclusion rule, or analysis will alter the findings.

These situations are not methodologically equivalent.

Changes made after researchers know the results create a greater risk that decisions will be influenced by whether those results are favorable, statistically significant, theoretically convenient, or otherwise desirable. SPIRIT guidance specifically warns that substantive amendments made partway through a study can complicate analysis and interpretation and may introduce bias when driven by trial data.

Changing the population changes more than recruitment

Suppose a study originally targets first-year and second-year university students. Recruitment among second-year students fails, and the researcher decides to proceed with first-year students only.

That is not simply a smaller sample. The population represented by the study has changed.

The researcher should revisit the research question, objectives, eligibility criteria, analysis, interpretation, and intended claims. If comparisons between year levels were part of the original design, those comparisons may no longer be possible.

The new boundary may be defensible, but it should be treated as a revised boundary rather than retrospectively described as an original delimitation deliberately established from the beginning.

Changing the sample size is not automatically changing the scope

Scope and sample size are related but distinct.

If a study still investigates the same population, question, variables, setting, and period but recruits fewer participants than intended, the substantive scope may remain unchanged. The problem may instead concern statistical precision, power, representativeness, information adequacy, or feasibility, depending on the methodology.

Conversely, reducing recruitment by dropping an entire population or setting can alter the scope even if the total sample remains large.

Do not use "scope change" as a catch-all label for every deviation from the research plan.

Changing an outcome or variable can be particularly consequential

Suppose a study originally identifies academic performance as its primary outcome. After seeing that the effect on academic performance is weak, the researchers notice a stronger effect on satisfaction and make satisfaction the centerpiece of the study.

That is very different from deciding before examining outcomes that academic performance cannot be measured reliably and formally revising the protocol.

Outcome-reporting guidance exists partly because selective changes to outcomes can bias the published evidence. CONSORT requires reporting changes to trial outcomes after commencement and their reasons, while empirical comparisons of protocols and publications have documented discrepancies involving omitted, downgraded, upgraded, or newly emphasized outcomes.

Changing a variable is not automatically inappropriate. Measurement failure, new scientific information, instrument availability, safety concerns, or other legitimate reasons may require revision. The important safeguards are timing, justification, approval where applicable, and transparent reporting.

Adding a research question can also expand the scope

Research often generates new questions. That is one of its attractions and, occasionally, one of its occupational hazards.

Discovering an interesting question during data collection does not necessarily mean it should become another primary objective of the ongoing study.

Ask whether the new question can be answered using the existing design without distorting the original inquiry. Does it require another population, instrument, method, outcome, or analytical strategy? Was the relevant evidence collected in a way suitable for answering it?

If the question is genuinely new, it may be better treated as exploratory or reserved for another study rather than retrospectively presented as a preplanned objective.

If new questions keep accumulating, the problem may instead be scope creep during the research process.

Expanding scope can be as consequential as narrowing it

Mid-study changes are often discussed as reductions caused by recruitment or resource problems, but scope can also expand.

A researcher may add another institution after recruitment begins, introduce a new participant population, collect an additional outcome, or add qualitative interviews to a quantitative study.

These additions may be scientifically useful. They can also change participant burden, consent information, privacy considerations, data-management requirements, analytical plans, sample-size needs, staffing, timelines, and the conceptual structure of the study.

Expansion therefore deserves the same methodological scrutiny as narrowing. More evidence is not automatically harmless evidence.

Human-participant research may require ethics approval before changes are implemented

If a study has research ethics committee or institutional review board approval, substantive changes may require an amendment before implementation. The precise requirements depend on the jurisdiction, institution, funder, study type, and applicable regulations.

SPIRIT 2025 recommends that trial protocols specify the process for amendments and how substantive changes will be communicated to relevant parties such as research ethics committees or institutional review boards, registries, and regulators. In regulated U.S. investigational-drug research, for example, significant protocol changes affecting safety, study scope, or scientific quality require formal protocol amendments, with specific approval requirements.

Do not assume that because a change seems methodologically sensible you are free to implement it immediately. Check the approved protocol and the requirements of the responsible ethics committee, institution, funder, registry, sponsor, or regulator before changing procedures.

Exceptions may exist for urgent changes needed to remove an immediate hazard to participants, but those are governed by specific rules rather than by general researcher discretion.

A scope change can affect informed consent

If a modification changes what participants will experience, what data will be collected, how those data will be used, the risks involved, or other information material to participation, the consent process may also need revision.

The implications depend on the nature of the change and applicable ethics requirements. Adding a new questionnaire may be minor in one context but substantial in another if it collects sensitive information. Adding a new procedure or substantially extending follow-up may require participants to receive additional information or provide renewed consent.

This is another reason substantive scope changes should be evaluated through the relevant ethics process rather than treated only as project-management decisions.

A scope change can make earlier and later data non-equivalent

Suppose a questionnaire is modified halfway through data collection. Participants recruited earlier answer one version; participants recruited later answer another.

Can their responses still be analyzed together?

Perhaps, but not automatically. The change may affect measurement comparability. Similar issues arise when eligibility criteria, intervention procedures, recruitment settings, instruments, observation periods, or data sources change during a study.

The researcher needs to consider whether evidence collected before and after the modification represents the same construct and study conditions sufficiently well for the intended analysis.

If not, the study may need separate analyses, sensitivity analyses, a revised estimand or interpretation, exclusion of particular evidence under defensible criteria, or another methodological response appropriate to the design.

Do not erase the original plan

One of the most important principles is surprisingly simple: preserve the history of the study.

If the original protocol included two populations and the revised study includes one, retain a record showing both versions. If an outcome changed, record the original outcome, revised outcome, timing, and rationale. If eligibility criteria changed, document what changed and when the new criteria took effect.

SPIRIT 2025 specifically recommends protocol version control using identifiers and dates, along with a record of amendments, to track protocol history.

This is not administrative fussiness for its own sake. Version history allows supervisors, collaborators, ethics committees, reviewers, and eventually readers to reconstruct how the final study emerged.

Prespecified and post hoc decisions should remain distinguishable

There is nothing inherently illegitimate about exploratory analysis. Problems arise when exploratory decisions are reported as though they were prespecified.

Suppose researchers discover an unexpected subgroup pattern and investigate it further. That can generate useful evidence and hypotheses. The transparent description is that the subgroup analysis was exploratory or post hoc if that is what occurred.

Likewise, if a new research question emerged during fieldwork, say so when that timing matters.

The purpose of distinguishing prespecified from post hoc decisions is not to forbid researchers from learning during research. It is to let readers evaluate the evidence with an accurate understanding of how the analyses and questions arose.

Qualitative research may involve planned methodological evolution

Not all methodologies treat the initial research plan as equally fixed.

Some qualitative approaches are intentionally iterative. Sampling, interview questions, analytical attention, or theoretical focus may evolve as researchers engage with the data. In grounded theory, for example, theoretical sampling is driven by emerging analysis rather than by a completely fixed participant plan established before data collection.

That flexibility should not be confused with undocumented improvisation. The methodological approach itself should justify the iterative process, and researchers should maintain an audit trail or other appropriate documentation showing how and why decisions evolved.

The key principle remains transparency, but what counts as an expected modification depends on the methodology.

Secondary-data research can also encounter scope changes

Scope problems are not confined to participant recruitment.

You may obtain a dataset and discover that an expected variable has extensive missingness, that particular years are absent, that one institution uses incompatible coding, or that the database does not represent the population you expected.

The temptation is to keep the original question and quietly analyze whatever remains.

A better response is to examine whether the available evidence still supports the original scope. If not, you may need to narrow the population, timeframe, variables, or question and explain why.

This situation deserves particular attention when available data force the scope to become narrower than originally intended.

A legitimate change can still create a limitation

Suppose recruitment difficulties require reducing a multi-institution study to one institution. The revised single-institution design may be coherent and methodologically defensible.

That does not mean the change has no consequences. The study may now provide less evidence about variation across institutional contexts than originally planned.

Similarly, replacing a measurement instrument because the original instrument becomes unavailable may be necessary while introducing comparability or validity concerns.

Researchers should distinguish the reason the change was justified from the limitations created by the change. Both can be reported without contradiction.

Some changes are large enough to create a different study

There is no universal threshold at which an amendment becomes a completely new project. But the possibility deserves attention when several core dimensions change simultaneously.

Imagine a study that begins as a quantitative survey of undergraduate students at three universities examining AI use and academic performance. It later becomes an interview study of faculty members at one university examining attitudes toward AI policy.

The topic remains generative AI in higher education, but almost everything that operationally defines the inquiry has changed: population, method, setting, phenomenon, evidence, and research question.

Calling this merely a "change in scope" may obscure the more important reality that the researcher is now conducting a substantially different study.

When expansion or revision creates independent questions and methodological structures, consider whether one project has effectively become several studies.

What a Defensible Mid-Study Scope Change Looks Like

The change does not automatically invalidate the study. Its defensibility depends on whether the revised question remains worthwhile, whether the remaining design can answer it, whether required approvals are obtained, and whether the history of the modification is reported accurately.

Common Mistakes When Research Scope Changes Mid-Study

When Scope Must Change, Change the Study Transparently Rather Than Quietly

When an unexpected problem arises, first determine whether it genuinely changes the substantive inquiry. Then evaluate the consequences before modifying the study.

Keep versioned records of the original and revised plan, including what changed, when it changed, why it changed, who approved or authorized the change where applicable, and which aspects of the study were affected.

Then align the entire project with the revised scope. Do not change the population in the methods while leaving the old population in the research question, or change the primary outcome while retaining objectives written for the original outcome.

If the difficulty arises because the available dataset cannot support the original plan, focus specifically on how to narrow scope without overstating what the remaining data can answer.

Before Changing the Scope of an Ongoing Study, Check:

Frequently Asked Questions About Changing Research Scope

A Study Can Change Without Pretending It Never Changed

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.