Failure is the foundation of science
Amanda Gardiner reflects on why failure is an essential and valuable part of scientific research
I love being a PhD student – but I don’t think I’m good at it. Or, more accurately, I don’t know if I’m good at it. In my undergraduate studies back in the US, I was drowning in coursework, and constantly complaining about how stressed I was. Now, I’m reminiscing on how I was consistently getting feedback on how I was doing, what things I was getting right, and what I wasn’t understanding and needed help with.
In my PhD, I’m mostly unsupervised. It’s more like having a job than being a student. I get occasional bits of feedback when I can squeeze a meeting into my supervisor’s already packed calendar, and despite these meetings being helpful, I often come out of them feeling unfulfilled and restless. It wasn’t until recently, now that I’m almost halfway through my PhD, that I realised why: I am terrified of getting told that my work, everything I’m pouring myself into, is wrong. And whenever I don’t get told ‘good job’ or ‘that’s right’, I feel like I’ve failed.
“Graduate students are six times more likely to develop depression or anxiety compared to the general population”
The way STEM is taught in our undergraduate degrees inherently sets us up for failure in our postgraduate studies. We are taught the ‘truth’ which we must know and memorise and spit out on command, and we must know more of this ‘truth’ than any of our peers to outcompete them for the best grades and rankings. We are taught that failure is bad, and must be avoided at all costs. Textbooks in undergraduate studies, and even the peer-reviewed scientific papers which form the foundations of communication in science, present research as a linear trajectory from ‘I have this hypothesis’ to ‘I created this study to test this hypothesis’ to ‘I have pretty plots to support my hypothesis’.
Martin Schwartz pointed out that this does a disservice to us in graduate studies and our future careers, as real research is much messier than this satisfying linear pathway. We should expect to feel stupid and fail at things. For the first time in our academic careers, we are taking a foray into the unknown, researching things that have never been researched by anyone in the world before. It seems obvious, then, that we don’t know the answers to our own questions, and that when devising experiments to answer a question never answered before that we won’t get it right the first time … or the second. Or even the third.
“Universities need to implement training sessions about imposter syndrome and failure, and how to overcome them”
But I, and many of my peers, see this as a mark of failure and stupidity rather than progress. Up until this point in our careers we have been rewarded for getting things right on the first try, not for admitting ‘I don’t know’. This, among a myriad of other stressors, is taking its toll on us. Imposter syndrome is widespread, and appears to persist as we continue through our degrees, with additional accomplishments only aggravating the problem rather than addressing it. Graduate students are also six times more likely to develop clinical depression or anxiety in comparison to the general population, and anyone who is not white, cisgender, or straight will have even higher chances of experiencing mental health issues during their graduate studies due to discrimination.
Systemic changes need to be put in place at a cultural and institutional level to start addressing these issues. Supervisors should be trained on how to be supportive of their students and how to deliver constructive feedback, while acknowledging that failure is a normal part of the process. Universities need to implement training sessions about imposter syndrome and failure, and how to overcome them. Conference organisers, journals, and societies need to stop rewarding the ‘publish or perish’ mindset.
But as individuals – in addition to supporting these institutional changes – we must remember that failure, rejection, and feeling stupid don’t mean that we are doing something wrong, but instead are inevitable parts of research as we work toward success. To be a researcher doesn’t mean that we never fail; it instead means that we persist in the face of failure and feeling dumb. Part of our journey as graduate students is to figure out how we persist, whether it be taking time to process our emotions, talking to peers or a mentor, or daydreaming about opening a bakery.
Last year, my older sibling finished their PhD in molecular biology. During their public thesis presentation, they showed a video of several dozen photos of failed experiment attempts, which took two years to finally work. There are many ways these defeats and setbacks could have been explained (most not fit for writing), but instead they described it in a way that has stuck with me since: “This is why it’s called ‘research’. If it worked the first time, it would simply be called ‘search’.”
Comment / Top of the slops: the competitiveness of college dining4 June 2026
Comment / The Cambridge drift1 June 2026
News / Cambridge researchers produce ‘world-first’ AI vaccine6 June 2026
News / Two more women allege sexual harassment by classics professor Simon Goldhill 28 April 2026
Features / One big happy (college) family?4 June 2026









