The Problem Isn’t the Problem: Why Researchers Need to Learn how to Solve Things
The experiment has failed again.
The results are not what you expected. A dataset has a gap you did not anticipate. A participant has dropped out. The equipment is not cooperating. A reviewer has questioned an assumption you were certain about. Somewhere between the research question and the answer, something has gone wrong.
For an early-stage researcher, this can be an uncomfortable moment. There is often an expectation that, after enough reading, training and preparation, you should know what to do next. But research rarely works that way. In reality, much of research involves dealing with things that do not go according to plan.
That is why problem-solving is much more than a useful extra skill. It is part of what makes research possible. Researchers develop and implement solutions to practical, operational and conceptual problems, often in situations where there is no obvious answer and where the problem itself may change as new evidence emerges.
And the process does not always begin with solving anything. Sometimes, it begins with asking a better question.
First, understand the problem
When something goes wrong, our first instinct is usually to fix it. But moving too quickly can mean solving the wrong problem. According to the paper “Teaching troubleshooting skills to graduate students”, researchers routinely encounter experiments, equipment and results that do not behave as expected, making the ability to diagnose problems and find solutions an essential part of research.
Imagine that a researcher is struggling to recruit participants for a study. The immediate question might be: “How can I get more people to participate?” That could lead to more emails, different recruitment channels or additional reminders. But perhaps the real issue is that the survey takes 40 minutes to complete. Or the invitation does not clearly explain why participation matters. Or the people being approached are not the right audience. The solution changes once the problem is understood differently.
This is one of the most useful habits an early-stage researcher can develop: before looking for an answer, spend time defining the problem. What exactly is happening? What do you know? What are you assuming? What information is missing? Is this actually one problem, or several smaller ones?
This kind of inquiry is closely connected to research itself. Asking questions about the themes, assumptions and conditions surrounding your own research can reveal problems that are not immediately visible. Sometimes, the most useful thing you can do is stop trying to solve the problem as it was originally presented and investigate what is really going on.
Turn uncertainty into something you can test
Once a problem has been defined, the next step is not necessarily to find a solution immediately. Often, it is to develop an explanation.
Suppose your results contradict your expectations. You might think the experiment failed. But there are several possible explanations. Perhaps the sample was too small. Perhaps a methodological choice introduced unexpected variation. Perhaps an assumption behind the analysis does not hold. Or perhaps the result is pointing towards something you had not considered. These possibilities can become hypotheses.
A hypothesis does not have to be complicated. It is a way of turning uncertainty into something that can be examined. You can ask what evidence would support it, what evidence would challenge it and what you could do to distinguish between different explanations.
This is where problem-solving and research methodology meet. You are not simply reacting to an unexpected result. You are using evidence to work out what might explain it and deciding how to test that explanation.
The important part is to resist the temptation to fall in love with your first answer. Good problem-solving leaves room for alternatives.
Test the solution - and be prepared to change it
Research problems rarely disappear after one attempt. You try a method. It does not work. You modify it. The new approach solves one problem but creates another. You consult a colleague, test another explanation and discover that your original assumption was only partly correct.
This can feel like going backwards. It is not necessarily. A solution is useful only if it actually addresses the problem. That means researchers need to assess the effectiveness of their own approaches rather than assuming that a solution is successful simply because it has been implemented.
The same applies when working with other researchers. Someone else’s approach may be useful, but it should still be examined critically. What worked in their context? Does the evidence support applying it to yours? What are its limitations? Could another approach work better?
This is an important shift from simply collecting methods to developing research judgement. And sometimes the assessment leads to an uncomfortable conclusion: the solution did not work.
That is still information.
A failed approach can rule out one explanation, reveal a limitation in the research design or point towards a different way of understanding the problem. What matters is what you do with the evidence.
When the evidence challenges the question
One of the more difficult steps in problem-solving is accepting that the original hypothesis may be wrong. Researchers spend considerable time developing hypotheses, building arguments and designing studies around them. It can therefore be tempting to treat unexpected evidence as something that needs to be explained away.
But evidence can do something much more valuable: change the question.
If the evidence repeatedly challenges an existing hypothesis, the response does not have to be to defend it at all costs. You can revisit the assumptions behind it, formulate a new hypothesis and test that instead. This is where problem-solving moves beyond fixing an immediate difficulty. Researchers begin to challenge existing explanations and develop new ones based on evidence.
Consider a researcher who starts with the assumption that a particular intervention should improve an outcome. The data show little effect. Rather than simply asking how to make the intervention work better, the researcher might ask whether the assumed mechanism was correct in the first place.
Perhaps another factor matters more. Perhaps the effect only appears under certain conditions. Perhaps the original concept needs to be reconsidered. Unexpected evidence is not always an obstacle between you and the answer. Sometimes it is the thing that takes you to a better question.
Some problems cannot be solved from one discipline
As research becomes more complex, the problems themselves often become less tidy.
Climate change, antimicrobial resistance, sustainable food systems, public health, digital transformation and the future of work are not problems that fit neatly inside a single academic discipline. They involve different types of evidence, competing priorities and perspectives from researchers, policymakers, industry, communities and other stakeholders.
For early-stage researchers, this means learning to recognise when a problem requires more than the tools of their own field. An environmental researcher may need to understand economic behaviour. A computer scientist working on healthcare may need to engage with ethics and clinical practice. A social scientist studying technology may need to understand how the technology itself works.
Interdisciplinary problem-solving does not mean becoming an expert in everything. It means recognising the limits of your own perspective and being able to work with people who bring different knowledge to the problem. That can change the solution entirely.
A problem that looks technical from one perspective may turn out to be social, organisational or ethical. A solution that works in a laboratory may fail when introduced into a real-world setting. Understanding those connections is part of tackling the complex problems that increasingly define research.
The bigger leap: solving problems that do not yet have answers
There is also a point where problem-solving stops being about responding to a problem and starts becoming about identifying one. Some of the most important research begins with a researcher noticing that something does not quite make sense.
Why does an established explanation fail in a particular context? Why does a widely used method produce inconsistent results? What assumption has everyone accepted without testing? What happens if we look at the problem from another discipline, another population or another scale?
These questions can lead to new hypotheses, new research projects and eventually new knowledge. This is one of the reasons problem-solving is not simply a practical skill for getting through difficult moments in a PhD. At more advanced levels, researchers are expected to challenge traditional thinking, identify gaps in existing knowledge and develop research that addresses problems in new ways.
The researcher is no longer simply asking, “How can I solve this problem?”
They are asking:
“Is this really the right problem to solve?”
And sometimes:
“What problem should we be solving next?”
Becoming comfortable with uncertainty
No researcher reaches a point where everything works exactly as planned. The further you go in research, the more likely you are to encounter questions without obvious answers, evidence that challenges your assumptions and problems that require expertise beyond your own.
The goal is not to eliminate uncertainty. It is to become better at working with it. For an early-stage researcher, that means developing a habit of moving from uncertainty to inquiry, from inquiry to hypotheses, and from hypotheses to evidence. It means testing solutions rather than assuming they work, learning from approaches that fail, considering perspectives beyond your own and being willing to change direction when the evidence demands it.
It also means recognising that problem-solving is not a single moment of inspiration. It is a process of questioning, testing, evaluating and adapting.
The next time your research does not go according to plan, resist the urge to ask only, “How do I fix this?” Start with a more useful set of questions: What exactly is the problem? What do I know? What am I assuming? What else could explain it? How could I test those possibilities? Is my proposed solution actually working? And what might the evidence be telling me that I have not considered yet?
Those questions can take you from solving an immediate research problem to developing something much more valuable: the ability to recognise, investigate and tackle increasingly complex problems throughout your research career.
Problem-solving as a research career competence
Problem-solving is one of the competences described in the European Competence Framework for Researchers (ResearchComp). Its progression reflects the reality that researchers’ problem-solving skills develop over time: from investigating questions within their own research and developing simple hypotheses, through tackling complex and interdisciplinary problems, to challenging existing hypotheses, developing new ones based on evidence and contributing new knowledge to unresolved problems.
The SMART Researchers project supports this development by helping researchers build the transferable competences they need throughout their careers. Through its learning and development activities, the project connects research skills with the wider capabilities researchers need to navigate increasingly complex research environments and work across sectors and disciplines.
Because research is not simply about finding answers to questions that are already well understood. Sometimes, it is about discovering that the question needs to change – and having the skills to work out what should come next.