The Role of Failure in Science: What Plant Researchers Learn from Experiments That Don’t Work
What does it mean to fail in science?
In scientific research, what we call “failure” does not always mean the same thing. While some failures arise from technical issues, others stem from the scientific process itself. In the first case, experimental problems due to procedural mistakes, inadequate methodologies, or suboptimal experimental design can often be prevented through experience and protocol optimization. In contrast, when a technically successful experiment reveals that the original hypothesis was incorrect, incomplete, or too simplistic to explain the biological phenomenon under study, failure is not a sign of poor research, but rather an opportunity to formulate better questions. After all, science itself is founded on the idea that we don’t know something. Getting everything right on the first try would, in many ways, defeat the point of pursuing research, and results that conflict with one’s original hypothesis are still valuable results.
Our experiences with failure
Flavia: As a plant scientist, I’ve made plenty of mistakes. One of them was at the beginning of my PhD, when I just couldn’t manage to extract plant DNA with decent quality and concentration. I repeated the protocol obsessively, probably a hundred times or so it felt, and still nothing. Then my PhD supervisor increased the centrifuge speed in one of the final steps, far beyond what I had been setting it to. Lo and behold, I finally got it.
What was my mistake? On one hand, I followed the instructions “to the letter”, not realizing that the protocol was designed for extracting DNA from much larger amounts of tissue (and bigger tubes), so I wasn’t considering differences in rotor type, tube size, or the height of the liquid column. On the other hand, I waited too long to ask for help, simply because I was embarrassed to admit I was struggling.
I’ve also faced other kinds of failures. One of the hardest was having to accept that, despite working with second-generation transgenic plants, I wasn’t going to find any CRISPR edits. The likely reason, which I realized too late, was that the transformation vector I was using carried Cas9 under a promoter that doesn’t perform well in the species I was trying to edit, combined with the fact that the type of edit I was attempting inherently has a low success rate. On top of that, the COVID-19 lockdown in 2019 prevented me from going to the lab to transfer many regenerating plants to the greenhouse, causing me to lose a significant amount of material as well. Lessons learned: first, each species has its own characteristics so not everything works everywhere. Second, it is important to regulate excessive optimism. Finally, I learned that science requires accepting the unexpected as an inherent part of the discovery process.
Cael: I certainly have my own collection of lab failures and lessons learned from them. Some were bigger deals than others: tripping after a full day in the lab and dropping half my DNA extractions was a pretty big (and embarrassing) one, while being unable to get a decent concentration on one out of dozens of samples was easier to get over. However, I’ve found that failures in other areas often provide even more meaningful, if difficult, lessons. The first time I received a rejection for a substantial grant felt like a real blow, and there is no sugar coating being told “no” when seeking funding. However, that failure ultimately resulted in a thorough, peer-reviewed experimental design that later became the foundation of my thesis proposal.
Perhaps the steepest learning curve in plant science for me has been the one presented by bioinformatics, and it’s also provided one of the most refreshing perspectives on failure that I’ve encountered so far. Anyone who works with any kind of code will know that you never get an error-free run on the first try (or second, or third…). This can become disheartening very quickly, especially for those of us who don’t have a computer science background. A suggestion from a mentor to think of each new error as a success allowed me to reframe each of those individual “failures” as a step closer to success.
We all fail at some point, even if we don’t like to admit it. And if you work in plant research, it’s impossible not to have made mistakes. Plant science presents its own unique set of opportunities for failure—we have to wait for plants to grow and reproduce, we are limited by short field seasons at the whims of climate change, and we often have much less control over our experimental systems than researchers in some fields do. Our professional journey is inevitably marked by errors and failures, yet we rarely talk about them openly because failure remains heavily stigmatized in scientific culture (Nunes et al., 2021).
The criteria for defining success and failure depend on personal experiences. However, for researchers, dealing with experiments that do not work is part of everyday life, as is experiencing failure in a broader sense beyond experiments, including rejected papers, unsuccessful grant applications, and career setbacks, all of which shape the scientific journey. Also important to remember is that plant science is a diverse field pursued by diverse people, and problems of accessibility and equity can often feel like personal shortcomings. Requiring accommodations for physically strenuous fieldwork or being unable to travel for conferences, for example, is easy to internalize as one’s own failure rather than the failure of an institution to support its researchers, and a greater emphasis on accessibility across plant science could go a long way in mitigating these feelings of personal inadequacy that only hold us back.
The daunting fear of failure
Considering an experiment as a “failure” simply because it did not turn out as expected is to place our wishes ahead of reality. So-called “negative” results provide valuable information that forces us to acknowledge that, despite all the effort, time, and resources invested, we may need to step back and start again by redefining the hypothesis or the goal.
These are particularly daunting concerns for early-career scientists, who face high competitiveness and expectations of excellence in their career development. Academia rewards breakthroughs and new discoveries but often penalizes the mistakes and dead ends that are an inherent part of the discovery process. This dynamic not only is contradictory to the definition of science, but also creates an atmosphere of anxiety and perfectionism, where scientists feel pressured to perform as if they know everything, something fundamentally incongruous with scientific discovery. Early-career researchers in particular often feel that they need to produce a perfect, foolproof proposal that accounts for every possible outcome before they can start their experiments, and that simply isn’t feasible. No one goes into a scientific project anticipating everything that could possibly happen, and placing that expectation on ourselves only prevents us from doing (and enjoying) the actual research. Accepting the inevitability of mistakes and unknowns in advance lowers the mental activation energy that often paralyzes us at the beginning of a project.
Moreover, when research funding is limited, designing and carrying out experiments becomes more stressful, as researchers often feel pressure to obtain statistically significant results that may determine the progress of their careers. This happens because experiments that contradict our hypothesis are not publishable, and no one escapes the logic of “publish or perish.” The success or failure of experiments ends up affecting the success or failure of our professional growth.
Although there is a growing movement to publish negative results in dedicated journals to give visibility to valuable data, it cannot be denied that, historically and overwhelmingly, there has been a publication bias in which only “successful” outcomes are shared (Mehta, 2019). This promotes an unrealistic image of the scientific process, as if the path to discovery were linear, straightforward and flawless. This idealized image of scientific discovery exacerbates our impostor syndrome (Mosher, 2017) when we fail to obtain “winning” results on the first attempt or when our work does not match expectations.
Failure is a great teacher
With limited resources and high expectations, it’s natural to fear mistakes and failure. However, the only person who never fails is the one who stays still. The best thing we can take from error and failure is what we learn from them (Vallery, 2016). Negative results that remain buried in a lab notebook don’t reflect the effort behind them, and they represent a missed opportunity for other researchers to learn from our experience. Acknowledging and sharing our mistakes or negative results can accelerate scientific progress because they may reflect the need to optimize protocols, change tools, or consider additional variables, while also helping to eliminate incorrect explanations and strengthen the robustness of scientific knowledge.
Some ways to embrace mistakes could include creating a dedicated space to discuss insights from unsuccessful experiments in doctoral theses, publications, or seminars; including “failed” steps and how they were overcome in conference talks; and encouraging evaluation committees to actively recognize and value the publication of negative results, as well as a candidate’s resilience and capacity for learning.
Experience with mistakes can also foster understanding and serve as a leadership tool when guiding others in the learning process, promoting greater patience with different training paces and the challenges that arise along the way. In addition, this experience guides how to avoid similar issues in the future and develop strategies to manage potential setbacks.
Not today, but maybe someday
Another frustrating form of failure is when our work is “rejected”. It’s the big NO we receive when we submit a manuscript for publication, apply for a research fellowship, or present a proposal for funding. In these cases, success is not about avoiding mistakes, but about being resilient and persistent. It’s a matter of trying again and again until chances and efforts finally play in our favor (Tay, 2017). When faced with rejection, we can choose to improve our communication skills, recalibrate our goals, be more flexible, and look for alternatives without losing momentum. Building resilience is a long process, but with every attempt, rejection hurts a little less, especially when we understand that we are all engaged in a continuous learning journey.
Final thoughts
Failure is inherent to scientific research, an essential part of the trial-and-error process that helps researchers to develop new methodologies or skills, discard unproductive paths, and build solid scientific knowledge. It is up to us to see it either as an obstacle or as a guide for learning. Many of us fear acknowledging failure in our research because we conflate it with complacency or “lowering the bar,” but accepting that failures will occur does not mean giving up or lowering your standards.
The widespread dissemination of “only successful” results can create a distorted perception of scientific reality, in which positive outcomes appear to be the norm while failures remain largely invisible. This selective visibility contributes to an unrealistic image of the research process, reinforcing the idea that progress is linear and consistently successful. In many ways, this resembles the culture of perfection often projected on social media, where only polished achievements are shared while struggles, uncertainty, and failure are filtered out. As a result, researchers may experience impostor syndrome when comparing their own inevitably imperfect and nonlinear processes (which include setbacks, negative results, or long periods of uncertainty that are rarely reflected in the published literature), with these idealized representations of success.
Perhaps a more realistic scientific culture would be one in which failure is not treated as something to be stigmatized or viewed with embarrassment, but as an intrinsic and unavoidable part of the research process. When failure is associated with professional risk, it can discourage researchers from taking intellectual and methodological risks, leading to increasingly conservative approaches to science. In contrast, recognizing failure as a normal outcome of inquiry could help create an environment where uncertainty is tolerated, exploratory ideas are valued, and the pursuit of new knowledge is not constrained by the fear of not succeeding. In this sense, how we define, communicate, and respond to failure ultimately shapes not only individual scientific careers, but also the kind of science we are able to collectively produce.
References
Mehta, D. (2019). Publishing negative results could help science move forward. Nature. https://doi.org/10.1038/d41586-019-02960-3
Mosher, S. (2017). Combatting the impostor syndrome in academic science: You probably are as smart as they think. Plantae. https://plantae.org/combatting-the-impostor-syndrome-in-academic-science-you-probably-are-as-smart-as-they-think/
Nunes, K., et al. (2021) Science students’ perspectives on how to decrease the stigma of failure. FEBS Open Bio. https://doi.org/10.1002/2211-5463.13345
Tay, J. (2017). Learning from rejections. Science, 355(6331), 1342. https://doi.org/10.1126/science.355.6331.1342
Vallery, T. (2016). Growth from failure. Science, 353(6298), 514. https://doi.org/10.1126/science.353.6298.514
______________________________________________
About the Authors
Cael Dant
Cael is finishing up an MS in plant biology and conservation at Northwestern University and the Chicago Botanic Garden, and serves as a 2026 ASPB Plantae Fellow. They love anything and everything to do with carnivorous plants, and their thesis research focuses on the ecology and physiology of the North American pitcher plant Sarracenia purpurea. Before graduate school, Cael spent five years working in government and international relations in Japan and still works as a Japanese-English translator while remaining involved in public policy from the science side. In their spare time they enjoy hiking, drawing, making pottery, crocheting, and growing plants.
Flavia Darqui
Flavia is a biotechnology graduate with a PhD in Biological Sciences and a 2026 Plantae Fellow. Her research focuses on transformation and CRISPR-based gene editing of lettuce as a crop model system. At present, she is studying lettuce genes involved in abiotic stress responses by generating knockouts using CRISPR/Cas9.
