Why Learning From Failure Works Better When Others Fail

We’re told that experience is the best teacher. We’re reassured that we can learn from failure. In recent years, we’ve gone so far as to argue that we need to celebrate failure. Mantras like “Fail More” or “Fail Faster” adorn many office walls where our prized motivational posters used to hang. There are definitely positive lessons to be learned from failure, but new research suggests that the failure of others might be a better source of learning than our own short-comings or mis-steps.

Researchers led by Emory University’s Diwas KC recently examined the experiences of cardiovascular surgeons to uncover whether success or failure was the better teacher, and if so, whose failure held better lessons? The team analyzed data from 71 cardiothorarcic surgeons over ten years as they performed over 6,5000 minimally invasive coronary artery bypass grafts, a complicated and relatively new procedure at the time of the study. They examined the rates of successful and unsuccessful procedures and also the process by which the surgeons were learning and improving their performance. As they analyzed the data, they found something striking: Failure is the best teacher mostly when someone else has failed. Surgeons learned best from their own successes and the failure of others, their own failures were much harder to learn from.

One possible explanation for this is attribution theory, specifically self-serving bias and fundamental attribution error. Taken together, these psychological biases predict that individuals more often attribute their success to their own efforts and their failures to outside circumstances, while simultaneously doing the opposite for others – assuming others’ failures are caused by their efforts and their success’ by outside circumstances. In the case of the surgeons, their own failures made for difficult learning material, since self-serving bias made it more difficult for them to notice and correct their own faulty actions. The failures of others, however, created a perfect learning laboratory as they could observe actions similar to the ones they took, but free of the tendency to attribute the outcome to bad luck. Interestingly, as the surgeons became more confident in their own abilities, by accumulating successful operations, they also became more likely to reflect on their own failures and learn from them.

The results carry strong implications for leading organizations. While the recent efforts to view failure more positively are a good start, perhaps leaders need to spend more time reframing individual failures as positive learning experiences. Seeing the failures in others not only aids learning, but helps to make one’s own failures less appear less threatening. In addition, organizations can help increase the learning of their members by taking time to review and reflect upon successes. Most often, when a project or procedure is successful, individuals are moved on to new projects without the benefit of a “post-op” review. Establishing a “post-op” protocol for every failure AND every success might ensure that experience becomes the best teacher, regardless of whether that experience was a success or a failure.

[Editor’s Note: This post first appeared as “Why We Learn Best From Others’ Failure” on SmartBlog on Leadership.]


  1. says

    Interesting commentary. As I read through the shared research I couldn’t help but think of the role research plays in moving most progressive ventures ahead. The word “failure” just doesn’t sit well with my background in healthcare, leadership, business or anything else. Yes, there are learning curves but accepting failure as a positive attribute does not sit well for me.
    In business would it be okay to administer a performance assessment to someone and say that they failed at their task but it’s okay for the operation? No. An array of corrective systems would be outlined with a very short timeline and if “failure” continued the individual would be gone.
    The idea that failure in the medical field, especially surgical procedures, might be acceptable is dangerous and could prove very costly. Again, research promotes such activity in a controlled environment as it does with any research.
    Even something as non intrusive s the field of leadership development, we research nearly everyday before we introduce a concept or design a new system to implement.
    As such, “failure” will not be acceptable in my world. Learning and research in a controlled environment will continue to be the only acceptable forum for risk based learning.

    • says

      Greg, I can certainly understand your feelings regarding failure in a medical profession. However, even if considered unacceptable, some level of failure is still inevitable. What I admire about the medical profession is that, through post-mortems, etc, they have made it possible to learn through failure…which is a great way to turn a negative into a positive. Thanks for the comment.

  2. Conor Teegarden says

    I can certainly agree with Greg’s point that in some contexts failure is unacceptable. The clearest example that I can think of is in military aviation, but even in that environment, like you mentioned David, failure exists and it is in building mechanisms to accept and learn from those failures that we can increase the likelihood of success. Risk is inherent, and it often becomes a matter of comfort with risk more than our actual perceived gains and losses that impact our decisions. What surprises me is that those industries with the highest stakes, medical and military, often have the most structured and emphasized learning environments. Could that be a lesson itself?

    David, I did appreciate the recognition of bias as the filter on our learning. It brought up the question in my mind: if we are able to mitigate or reduce the impact of our bias, would we be able to learn even more from our own failures? While the research supports effective ways of gleaning learning by accepting our biases, I’d be interested in seeing what greater learning might be achieved if we can overcome them. Thanks for the article!

    • says

      Thanks for the comment Conor. I think any time you are aware of your own bias you have the opportunity to learn more about your own actions. There’s a good amount of research supporting the idea of “getting some distance” from yourself as an aid to decision making. I’m not aware of a specific study looking at that idea in the context of failure, but I’ll start looking for one.