Survivorship bias in science: is individual resilience the most important quality of a good scientist?

A damning piece was published in The Guardian claiming that 40 percent of academics are considering leaving their position due to the stress and strain of the academic career path. We’ve spent a considerable amount of time in this column talking about the leaky pipeline in academic science but rarely do we discuss those who have “made it”, and call it quits after doing so. This seems to indicate that the academic environment is retaining even fewer people in the long run and leaves one to wonder what factors are driving them away. It is clear from the article that academics are under stress and many are leaving, but why they leave and who ends up remaining have substantial implications for moving forward in how we educate and train scientists.

What keeps academics in academic science?

There is no doubt that the career of an academic scientist involves constant rejection – we learn to get rejected-over and over again in grant applications, paper reviews, etc., and some of them are really nasty. With all of the rejection, you need to build coping strategies. Sometimes the work or idea is indeed misunderstood and other times we delude ourselves with ever-more-elaborate reasons for the rejections. Either way, an academic career can sometimes feel like a constant barrage of obstacles and those operating within it need to hone their coping strategies. The work and careers of academics are under constant scrutiny, with evaluation exercises and coping strategies becoming so important that it begs the question of whether an academic scientist can survive without extreme resilience. This leads to the further, more dangerous, query of how this necessity for extreme resilience might position the future direction of academic science.

Coping strategies aren’t always nice

While it is all well and good to imagine healthy coping strategies like having a plan B, C, D, and E for where your rejected manuscript might go next or having planned downtime or re-focusing sessions, etc, the reality is that many people will find coping strategies that are less pure in their intention. Just like when athletes are faced with extreme competition, their responses will be quite varied – some will train harder, others adjust their expectations, and some go right ahead and break the rules. I remember distinctly when a postdoctoral colleague said to me “better to be doing trendy science that ticks the boxes than no science at all” (i.e., if you don’t play the game and compete well, you won’t have a job in academic science). While I disagree with the sentiment, I can see that many people are faced with the question of “how do I survive?” and we cannot always predict what form their reaction will take.

A telling sign that there is a wider spread problem is that Wellcome (one of the U.K.’s largest funders) has identified areas for significant improvement in research culture including creative collaborations, diversity and inclusion, and research honesty/integrity. None of these areas seems particularly compatible with the current system that valorises and rewards the individual scientist above all else.

Policy changes are needed, but who gets a say?

As a scientific community, we must think about the consequences of potential biases in the retention of scientists, but to do this effectively we should be careful not to let survivorship bias play into these policies as well. At present, it seems that those who make the rules are the survivors – they had it tough and came out stronger. This does not mean that everyone learns and grows in this way. More worrying is that not everyone who survives plays by the rules – the “old boys” network still exists in many spaces, data massaging and outright fabrication are becoming more common, and careers can be buried for completely non-scientific reasons.

One source of information that is rarely tapped into for improving the way we educate, train, and prepare young scientists is former academic scientists. If we could have exit interviews from a couple of hundred postdoctoral fellows or these crowned lab heads who opt to leave after reaching the top of the hill, how different would their advice be from the survivors’ advice that currently dominates institutional policy?

Survival via cooperation

An excellent editorial was published last year in Nature where it reminded me that while the selfish instinct may seem logical and rewarding in the face of competition, humans actually have a long history as a social species of finding their best success(es) through cooperation. From this article, I think there are some salient points worth reminding readers who would like to encourage more cooperation to consider. First, you need to find a mechanism to “signal your intent and good qualities to potential partners”; secondly, you have to see the long-term value for everyone finding success (i.e., feeding many potential hunters with a single killed animal makes sense for the next hunt) and third, you need to build trust between the cooperative parties (i.e., getting people to think about collective good over individual good as in a Prisoner’s Dilemma scenario).

Upon closing, I want to share a personal story about when I felt “happiest” as a scientist. During my doctoral training, many of my fellow trainees and I pulled long hours and we had late night and weekend meetings with our supervisor (sometimes at her house for an entire day). While some would view this as a horrible sweatshop of graduate student and postdoctoral fellow exploitation, I actually felt like we were part of a family with a common purpose and feel that this enabled me to work in a dedicated fashion without clocking hours in a traditional sense. Could other factors (i.e., living in Vancouver, no children or mortgage) have contributed? Completely possible, but I would still hope that letting people enjoy science and the company (and shared success!) of other scientists would be a good thing to strive for.