Research vs. Professional Practice, Evaluation, Audit, and Quality Improvement: Where Is the Boundary?

If You Collect and Analyze Data Systematically, Are You Automatically Doing Research?

A hospital measures infection rates before and after changing a procedure. A university surveys students to evaluate a new support program. A school reviews whether teachers are complying with an assessment policy. A clinician systematically examines patient outcomes to improve a local service.

These activities can look remarkably like research. They may involve research questions, surveys, interviews, routinely collected records, statistical analysis, qualitative analysis, comparison groups, and formal reports. Some may even eventually be published.

Yet professional practice, program evaluation, audit, and quality improvement are not automatically research simply because they collect and analyze data. The boundary often depends on the activity's primary purpose, the type of knowledge it is designed to produce, how the findings will be used, and the formal definitions governing the setting.

Similar Methods Can Serve Different Purposes

How Research Differs From Other Systematic Professional Activities

The presence of data does not settle the classification

Research is not the only activity that collects evidence systematically. Hospitals monitor outcomes. Universities evaluate programs. Businesses examine performance. Schools assess interventions. Government agencies conduct audits. Professionals review cases and outcomes to improve practice.

These activities may be methodologically sophisticated. They can use surveys, interviews, observations, databases, statistical models, qualitative analysis, or experimental-style comparisons.

The existence of systematic data collection therefore cannot, by itself, tell you whether an activity is research. You need to ask what the investigation is designed to accomplish.

Research primarily seeks to develop or contribute to knowledge

Research addresses questions whose answers are intended to contribute to a broader body of knowledge rather than solely resolve an immediate operational need.

In the U.S. federal human-subjects framework, research is defined for regulatory purposes as a systematic investigation designed to develop or contribute to generalizable knowledge. This is an operational definition for a particular regulatory system, not a universal philosophical definition of all research.

Still, the emphasis on knowledge contribution helps illuminate why research as systematic investigation differs from many routine professional activities.

Professional practice primarily aims to act competently in a particular situation

Professionals routinely gather and interpret evidence while doing their jobs. A physician examines symptoms and test results to diagnose a patient. A teacher reviews assessment results to decide what to reteach. An engineer investigates why equipment is malfunctioning. A counselor evaluates information to determine an appropriate course of action.

These activities can be systematic and evidence-informed without being research. Their primary purpose is usually to make a professional decision about a particular client, patient, student, system, or situation.

The distinction should not be interpreted as a hierarchy. Professional practice is not inferior research. It is a different activity with a different immediate purpose.

Evaluation asks how well a program, policy, or activity is working

Program evaluation systematically examines a program, policy, service, or organization to support judgments and decisions about it.

The CDC's current Program Evaluation Framework describes evaluation as a systematic activity used to understand and improve programs and organizations, assess effectiveness and efficiency, support accountability, and inform decisions. It explicitly distinguishes evaluation from research by purpose: research primarily aims to contribute to generalizable knowledge, whereas evaluation aims to improve programs and organizations and produce findings useful for decision-making.

An evaluation might ask whether a student-support program is reaching its intended participants, whether implementation follows the original plan, whether expected outcomes are occurring, why a program works differently across sites, or whether resources are being used efficiently.

These can be intellectually demanding questions. The distinction is not that evaluation is methodologically casual. Its primary orientation is toward judging and improving the evaluand, meaning the specific program, policy, service, or organization being examined.

Evaluation can use methods that are indistinguishable from research methods

An evaluator may administer validated questionnaires, conduct interviews, analyze administrative records, compare outcomes over time, use statistical models, or even construct a comparison group.

None of these techniques belongs exclusively to research.

This is why classifying an activity by its method can be misleading. A randomized or quasi-experimental design may be used in an evaluation. Interviews may appear in both qualitative research and service evaluation. Statistical significance tests may appear in either.

The method tells you how evidence is being examined. The purpose tells you why.

Audit asks whether practice meets an expected standard

Audit typically examines whether current practice, performance, documentation, or outcomes conform to predetermined criteria or standards.

A clinical audit, for example, might examine whether eligible patients receive a recommended procedure within a specified period. An academic audit might examine whether required assessment processes are being followed. A financial audit examines whether records and practices meet applicable standards and requirements.

The central question is often not "What new knowledge can we produce?" but "Are we doing what we are supposed to be doing, and where are the gaps?"

Audit findings may lead to corrective action followed by another audit to determine whether performance improved.

Quality improvement focuses on making a particular system work better

Quality improvement, commonly abbreviated as QI, generally involves systematic efforts to improve processes, services, safety, efficiency, or outcomes within a particular setting.

In health care, QI might involve introducing a revised handoff procedure, monitoring errors, studying workflow, testing process changes, and iteratively modifying the system.

OHRP guidance notes that quality-improvement activities limited to delivering health care and measuring or reporting performance for clinical, practical, or administrative purposes do not satisfy the HHS regulatory definition of research. It also recognizes that some QI projects can have a research purpose as well as an improvement purpose.

This is an important boundary: QI and research are not mutually exclusive simply because they have different typical purposes.

The categories overlap because one project can have more than one purpose

Suppose a hospital introduces a new medication-safety process because it wants to reduce local prescribing errors. Staff monitor error rates and adjust the process repeatedly. That activity may primarily be quality improvement.

Now suppose the team also designs a prospective comparative study to establish whether the intervention produces an effect that can inform practice beyond that hospital. The project may now contain a research component.

OHRP specifically recognizes this possibility. A QI project can constitute research when it includes a research purpose, such as systematically testing an intervention to establish scientific evidence about its effectiveness.

The correct question is therefore not always "Research or QI?" Sometimes it is "Which parts of this project are QI, which are research, and which requirements apply to each?"

“Generalizable knowledge” is useful but should not be used mechanically

The phrase generalizable knowledge is central to some regulatory definitions of research, but it can be misunderstood.

A study does not cease to be research merely because it occurs at one institution. Case studies, qualitative research, and context-specific investigations may still contribute to scholarly knowledge without making statistical population generalizations.

Conversely, an evaluation can produce lessons useful to other organizations without automatically becoming research under every regulatory framework.

The intended contribution, design, context, and governing definition all matter. Researchers should therefore use generalizable according to the authority whose definition is being applied rather than as a universal methodological slogan.

Local relevance does not make something non-research

Research can address highly local problems. A researcher may investigate a phenomenon within one university, hospital, community, organization, or classroom because that setting provides an informative case.

The question is not simply whether findings are useful locally. Research findings often should be useful to the setting being studied.

What matters is whether the investigation is designed primarily to answer a research question and contribute to knowledge, rather than solely to manage or improve the immediate service.

Publication intent does not determine whether something is research

A particularly persistent misconception is that an activity becomes research when the team decides to publish it.

OHRP explicitly states that intent to publish is insufficient for deciding whether a quality-improvement activity fits the regulatory definition of research. Nonresearch activities may be published because their experiences are useful to others, while research can remain unpublished.

This follows the broader distinction between conducting research and publishing research. Publication is a communication process, not a magic conversion mechanism.

Ethics review should not be decided by relabeling the project

Calling a project "quality improvement," "evaluation," or "audit" does not automatically exempt it from research requirements if its actual purpose and design meet the relevant definition of research.

The reverse is equally important. Not every systematic evaluation involving people automatically requires the same review process as non-exempt human-subjects research.

Ethical responsibilities can exist even when an activity is not formally research

A project falling outside a particular research regulation is not thereby ethically unrestricted.

Quality improvement, evaluation, audit, and professional practice may still involve sensitive personal information, confidentiality obligations, informed participation, fairness concerns, professional standards, data governance, institutional policy, or other ethical responsibilities.

Research ethics regulations establish particular protections for activities within their scope. They do not define the entirety of ethical professional conduct.

A useful distinction is purpose, but purpose is not the only consideration

These descriptions are useful starting points, not universal legal definitions. Fields and jurisdictions use the terms differently, and hybrid projects are common.

The same dataset can support both operational work and research

An institution may collect student records to operate its academic programs. Administrators may analyze those records to monitor retention. Evaluators may use them to assess a particular intervention. Researchers may later analyze the same records to answer a scholarly question.

The data do not determine the classification by themselves. Purpose, design, governance, consent or legal basis, identifiability, and applicable institutional requirements may change when the same information is used for another activity.

This is another reason research using existing data requires attention to how and why the data are being reused.

The Same Student-Success Problem Can Produce Four Different Activities

The visible activities may overlap considerably. Students may complete the same questionnaire, and administrative records may contribute to several analyses. Classification depends on the purpose and design attached to those uses, together with the formal rules governing the project.

Common Misconceptions About the Boundary of Research

Classify the Purpose Before Choosing the Label

If your project sits near the boundary, write down what you intend to accomplish before deciding what to call it. A project description often reveals more than the label attached to the proposal form.

When the classification affects formal obligations, the safest methodological skill is knowing when classification is not yours to decide alone.

Before Deciding Whether Your Project Is Research

Frequently Asked Questions About Research, Evaluation, Audit, and Quality Improvement

The Boundary Depends More on Purpose Than on the Methods You Use

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.