03 · What You Need to Know
Research Can Contribute Without Producing a Solution
A research problem is not necessarily a problem that needs fixing
The word problem causes much of the confusion.
In ordinary language, a problem is usually an undesirable condition that should be corrected: students are dropping out, a treatment is ineffective, a system is inefficient, misinformation is spreading, or an organization is losing employees. Research addressing such conditions may indeed seek knowledge that contributes to a solution.
In scholarly work, however, a research problem can also be an intellectual problem. Something important is unknown, inadequately explained, disputed, inconsistently observed, poorly measured, or theoretically unresolved.
Practical problem
A condition in the world that someone may want to improve, prevent, change, manage, or address.
Research problem
An important uncertainty, limitation, contradiction, or unresolved issue in what is known and how it is known.
The two can overlap, but they are not identical. A practical problem may contain several research problems. Conversely, a research problem may be scientifically important even when nothing is obviously “wrong” in the practical sense.
Description can be a contribution
Before researchers can explain or change a phenomenon, they may need to establish what is actually happening.
A descriptive study might estimate prevalence, document behaviors, map variation across settings, characterize a population, identify patterns over time, or establish features of a phenomenon that have not been adequately documented.
Consider an emerging technology used informally by university students. Before asking whether the technology improves learning, researchers may need credible evidence about who uses it, for which tasks, how frequently, under what conditions, and with what patterns of use.
That study does not solve the educational issue. It may not even determine whether the technology is beneficial. Yet reliable description can replace speculation with evidence and expose more precise questions for subsequent investigation.
Watch Out
“Nobody has described this before” is not sufficient justification by itself. Description becomes a meaningful contribution when the phenomenon is consequential enough that knowing its distribution, characteristics, frequency, context, or variation improves understanding or enables important subsequent work.
Explanation can matter even when it does not tell anyone what to do
Some studies seek to understand why something occurs rather than how to stop, improve, or manipulate it.
Researchers might investigate why students respond differently to feedback, why an association varies between contexts, how people form judgments about information credibility, or what mechanisms could account for an observed pattern.
An improved explanation can change how a field understands a phenomenon. It can distinguish between competing accounts, reveal previously overlooked mechanisms, identify boundary conditions, or generate predictions that later research can test.
None of those outcomes necessarily provides an immediate solution. Their contribution lies in making the phenomenon more intelligible.
Testing an assumption can be valuable even if the assumption survives
Research communities operate with assumptions. Some are strongly supported. Others become conventional because they have been repeated frequently, worked reasonably well in familiar settings, or simply escaped serious scrutiny.
A study can contribute by subjecting an important assumption to a credible test.
If the assumption is supported, the contribution may be stronger evidence about its robustness or scope. If it is challenged, the findings may prompt researchers to reconsider theories, measurements, methods, or interpretations built upon it.
The value therefore does not depend entirely on obtaining a surprising result. A well-motivated test can be informative because uncertainty existed before the test was conducted.
This becomes especially important when deciding whether to avoid a research question because the expected result seems obvious. What seems obvious is not always the same as what has been adequately established.
Research can clarify where a theory works and where it does not
Theoretical contribution does not always require inventing a new theory. Research may contribute by examining whether an existing explanation holds under conditions where its applicability is uncertain.
Suppose a theory developed largely from studies of face-to-face learning is increasingly used to explain behavior in asynchronous online environments. A study could examine whether the relationships predicted by that theory remain evident in the new context.
If they do, the study provides evidence about the theory's applicability beyond the settings in which it has usually been examined. If they do not, the result may reveal boundary conditions or suggest that the explanation needs refinement.
Neither outcome “solves” online learning. Both may improve the theoretical basis on which later research proceeds.
Methodological improvement can be a contribution in its own right
Sometimes the most important result of a study is not a new substantive finding but a better way of producing evidence.
Researchers may develop or validate an instrument, compare analytical approaches, improve a coding framework, examine measurement invariance, establish the reliability of a procedure, create a reusable dataset, or identify biases in an established method.
Imagine that researchers frequently compare two groups using a particular scale. A methodological study finds that several items function differently between those groups. The study has not solved the substantive issue researchers were originally interested in. Instead, it has revealed that some existing comparisons may not mean what researchers assumed they meant.
That can be consequential because better methods affect the credibility of many later findings.
Research can tell us that the evidence is weaker than we thought
Not all useful research increases certainty. Occasionally, good research reveals that previous confidence was unjustified.
A replication may fail to reproduce an influential finding. A systematic investigation may show substantial heterogeneity between contexts. Better measurement may weaken an apparent association. A sensitivity analysis may demonstrate that conclusions depend heavily on assumptions.
These findings can feel less satisfying than a clean solution, but they perform an important scientific function: they recalibrate what researchers are entitled to claim.
Knowing that an answer remains uncertain is not the same as knowing nothing. It can prevent premature conclusions and identify where additional evidence is genuinely needed.
A null result can still answer an important question
Suppose researchers compare two instructional approaches because there are credible reasons to expect one to outperform the other. A rigorous study finds little evidence of a meaningful difference under the conditions examined.
The study has not produced a superior intervention. It may nevertheless have answered an important comparative question.
Its value depends on the design, precision, prior uncertainty, and interpretation. Failure to reject a null hypothesis does not automatically establish equivalence or prove that “there is no difference.” Still, appropriately designed research can provide evidence that an expected advantage is smaller, less consistent, or more context-dependent than assumed.
The question of whether it is worth conducting research when you expect no difference therefore cannot be answered merely by asking whether the study will discover a solution.
Finding that “nothing happens” may itself be informative
The same logic applies beyond statistical comparisons. Researchers sometimes investigate whether an event, intervention, exposure, policy, or condition produces an expected change and discover that the anticipated effect is not evident.
If the expectation was theoretically or practically consequential, that absence can matter.
It may challenge a proposed mechanism, weaken the rationale for an intervention, suggest that contextual conditions matter more than expected, or redirect attention toward alternative explanations.
The relevant issue is whether the absence was investigated with a design capable of detecting the phenomenon of interest. A poorly powered or poorly measured study cannot transform lack of evidence into a meaningful discovery simply by calling it a null result.
When the possibility of no observable effect is central to the question, it is worth considering separately whether the answer might genuinely be “nothing happens”.
Research can create something other researchers need
A contribution does not have to be a conclusion. Research can produce resources that make subsequent inquiry possible or better.
These may include datasets, corpora, archives, instruments, protocols, taxonomies, software, validated measures, conceptual frameworks, or methodological procedures.
The value of such outputs depends on their quality and usefulness. A dataset is not important merely because it exists. But a carefully constructed resource addressing a genuine limitation in available evidence may support questions far beyond those examined in the original project.
This is one reason it helps to ask who needs the answer to your research question. Sometimes the immediate users are other researchers who need better evidence, methods, concepts, or resources before a practical solution can reasonably be pursued.
Research value can extend beyond direct problem-solving
Research funders and assessment systems themselves often distinguish several forms of contribution. The U.S. National Science Foundation, for example, evaluates both intellectual merit, which concerns the potential to advance knowledge, and broader impacts, which concern potential benefits to society. Its guidance also recognizes that fundamental research may take years to produce transformative outcomes and that specific eventual outcomes can be difficult to predict.
Similarly, UK Research and Innovation describes research impact broadly. Its Engineering and Physical Sciences Research Council notes that impact can result from advances in knowledge, understanding, methodology, theory, and application, as well as from products, processes, services, and knowledge exchange.
These frameworks differ in purpose and should not be treated as universal definitions of research value. They do, however, illustrate an important point: scholarly contribution is broader than producing an immediate fix.
Not solving a problem does not exempt a study from demonstrating significance
There is an opposite mistake worth avoiding. Once researchers recognize that studies need not solve practical problems, it can become tempting to treat intellectual curiosity as sufficient justification for anything.
It is not.
A question can be unanswered yet trivial. A new dataset can duplicate resources that are already adequate. A theoretical extension can add almost nothing to existing explanation. A descriptive study can document a pattern nobody has a compelling reason to know.
The issue returns to the difference between an interesting and an important research question. Research does not have to solve something, but it should change something meaningful about the state of knowledge or our capacity to investigate, understand, or make informed judgments about the phenomenon.
Contribution is often cumulative rather than dramatic
Researchers sometimes imagine contribution in excessively large terms. A study must supposedly transform a field, settle a debate, solve a longstanding problem, or produce an entirely new theory.
Most research does not work that way.
Knowledge is often cumulative. One study estimates a relationship more precisely. Another examines a different context. Another identifies an exception. Another improves measurement. Individually, these contributions may be modest. Together, they can substantially change what a field knows.
The important question is therefore not whether your study changes everything. It is whether it makes a contribution that is sufficiently meaningful relative to the question, evidence, existing literature, and resources required to produce it.
A contribution can be small without being trivial, a distinction examined more closely in the guide on whether a very small research contribution can still be worth making.
