Hello, my fellow PIs, here is a question for you: Did you get trained well for your job?

Silly question, of course you did. Years of study and examinations culminating in a PhD have obviously trained you well in all things science.

But that’s not what I mean. Details of experiments and algorithms –what you learn in a PhD– are only a small part of a PI’s job. Once you start leading a group, the tough nuts to crack are people-problems.

My own job title is ‘group leader’ which is a pretty appropriate description of what I do. Much more appropriate than Principal Investigator, because I actually don’t investigate that much myself. Either I propose what others should investigate in the future (in my grants) or I describe what others have investigated in the past (in my talks). Their job is investigating, my job is leading. But when I started my job, nobody had really prepared me for that.

And I am not an exception. There is a general lack of formal leadership training and support for PIs. Most of us get trained in an apprenticeship-like system, where we observe our advisors during our PhD and postdoc stages and hope that we can pick up some useful skills. And after that it’s “learning on the job”, a process of trial and error in which you figure out how to run a lab while running one.

A lab is more than just a PI

This lack of leadership training can lead to a very biased view of what it means to be a PI. The problem is perfectly exemplified by a flyer from the University of Cambridge Personal and Professional Development department. (A link to the online resource is here, but you need a Cambridge login)

A page called ‘leadership essentials’ lists quotes from people throughout the university on how they understand their role. One of these quotes by a Wellcome Trust Research Fellow caught my attention:

The key responsibility for being a PI running your own lab is to get the lab doing really well scientifically, producing good work and good publications, because that’s good for everybody. [my emphasis]

I certainly agree that producing good papers is *a* key responsibility. Without good science you soon won’t have neither funding nor a lab.

For me the quote perfectly describes the perspective of a PI: The sun shines, if my lab publishes in Nature every year. Department heads support this attitude because visible scientific output measurable by impact points looks good on their annual reports.

What I disagree with is the end of the quote: “… because that’s good for everybody“. To me this sounds naive. What is good for the lab head  is not necessarily good for all lab members.

Why? Simply, because the PI perspective is different from the student and postdoc perspective. For a successful career, students (and postdocs) need to maintain “an intact, enthusiastic spirit of curiosity, unbroken by terrible experiences*” – and this is not necessarily identical with publication output.

I have heard too many stories of students, who left science even with a first-author Nature/Science/Cell paper because of frequent frustrating and terrible experiences. Some felt micro-managed or bullied by their supervisors, some felt completely alone and unguided.

The biggest cause of frustration I hear from PhD students is bad supervision. And I don’t speak of weak students who need spoon feeding, but strong students whose initiative and creativity are stifled by their supervisor’s micromanagement or lack of attention.

And the same can be true for postdocs: Just think of big labs with dozens of postdocs some famous people have. This big workforce is good for the PI, because more workers means more papers. But all these postdocs compete for the PI’s attention in a Darwinian struggle for career survival. A lucky few win, most lose – no matter how talented they are.

So, in summary, there are strategies for a PI to maximize their research output that are exploitative and not good at all for the people working with them.

(I am not implying the WT fellow quoted is doing any of this. The single sentence I quoted is taken out of context and will certainly not encompass the fellow’s complete leadership style.)

What is scientific leadership?

The problem is the common definition of scientific leadership as “Big grants, big papers, big vision”! That’s all true, but this definition overlooks the most important type of scientific leadership: the leadership within your own lab.

At my institute in Cambridge we now ask leadership question in job interviews. Something like ‘What was the worst situation you had with an advisor or advisee and how did you handle it?’ But no one had asked me any leadership questions in any job interview anywhere. They all just asked ‘what is the title of your first grant’. My papers counted, my abilities as a supervisor didn’t.

This problem is aggravated by how isolated scientific labs are even in these modern days of global collaborations. Each lab is an island. PIs are independent researchers and what happens in their lab stays in their lab. They rarely get any feedback on their performance within the lab.

There are of course mechanisms to handle the big drama, when a supervisor-student relationship breaks down completely, but how PIs handle the little daily drama that comes with being responsible for other people is completely up to them — mostly without any guidance or feedback. If they don’t proactively and consistently ask for feedback, they might not even know what the people in their own lab are thinking.

Continuous leadership support

Many universities have postdoc development programs, but I don’t know of any PI development programs. There are generally few support networks for PIs. The only exception I know is the Weizmann’s Young PI forum and their webpage hasn’t been updated in years – I wonder if they are still active.

I believe most departments are failing in their responsibility to develop and support PIs. To change this, I will outline here key elements of a scientific leadership program for faculty of all ages.

1. In the beginning

A great start are EMBO lab management courses for postdocs and newbie PIs (and other similar workshops worldwide on learning to lead a lab).

These courses teach you about personality and leadership styles and make you aware how your own way of doing things might be experienced by other people who work differently (but not necessarily worse). This is an essential step to break out of the self-focussed leadership style of ‘everything’s good when I’m happy’. The courses also teach you coaching skills and techniques to handle conflict in the lab. Pretty much an essential survival kit for your first days, months, years as a PI.

In most universities participation in such a course is not mandatory. This can be an issue, because the self-selected group of people who go there voluntarily are the ones who are already aware of how important leadership skills are. The problem are the people who hide their lack of people and leadership skills behind ‘being too busy doing great science and writing important papers’.

Department heads, this is to you: Make attending a leadership course mandatory for new faculty! Regardless of how much they are able to take out of it, you have made a strong point how important leadership skills are in your department.

2. Continuously throughout a career

Some kinds of continuous support are well established. In many places you get feedback on grant proposals by senior colleagues and can rehearse important talks in front of other faculty members. This is all good, but not what I am interested in here. Rather, who do you turn to if you have a conflict in the lab? When you don’t know how to handle a difficult situation or a difficult collaboration partner?

Ideally you can talk to your peers in an environment you feel secure in. I already mentioned the Weizmann’s Young PI forum. Milo and Schuldiner’s paper describes how they set it up and topics they work on, e.g. dealing with rejection and the imposter syndrome we all seem to suffer from. ISCB is building something similar with their Young PI meetings at ISMB each year.

You need a minimum size to keep a peer network running. At the Weizmann they often had 15 young faculty in the room. This might be too big for many departments, but could be set up in an university-wide context, like across all biological sciences.

While your peers are all at a similar stage of their careers, you will also need some support from mentors, scientist who are ahead of you and can guide you with their experience. Be careful how you choose them, if they are too close to you they can’t guide you and if they are too far away they can’t understand your struggles. I have fared well engaging with scientists, who are 5-10 years ahead of me, at important steps of my career.

While relationships to mentors can be quite informal, there is also an opportunity to hire professional coaches. Career coaching for scientist seems to be an up-and-coming market.

[V]olunteer mentors are only able to go so far with their time and advice. A professional coach is paid to carry you through to a successful outcome, and the good ones take this responsibility seriously.*

The experience we had at my institute working with a professional coach shows that this is completely true.

Department heads, this is to you again: put a small budget aside for coaching your faculty. Compared to the millions you spend in your research, a tiny investment will earn you lots of benefits by supporting faculty at all career stages.

3. For tenure decisions

When I had my tenure review in 2014, it was a top-down procedure. The institute director organized a panel of senior scientists to come to Cambridge and interview me, as well as a second group to write letters about me. My group and my peers were not involved in the process at all.

What a wasted opportunity! Look what the tenure panel could have learned from talking to …

.. my students or postdocs. Is your boss a good supervisor to his students? Does he exploit his postdocs to further his own career? Does he succeed in establishing a nurturing environment in his group? Is he accessible? Is he a micromanaging tyrant who bullies poor innocent students into submission until they cry?

… my collaboration partners. How is it, working with Florian? Is he delivering what he promises? Is he fair and generous when discussing contributions? Or is he a cut-throat careerist who will fight tooth and claw for a better position on a paper?

… the other group leaders at the institute. Is Florian an arrogant jerk that nobody wants to talk to? Or is he a good citizen who contributes to the collaborative atmosphere at the institute?

Nobody asked these questions, yet none of them can be answered by a CV and publication list. Nobody asked these questions, yet all of them are important if you have to decide whether to keep someone in a department or not.

Most department heads will know the answers to some of these questions. But –and that’s the key point– these answers were not an explicit part of the formal review process, and they should have been.

4. At regular intervals

What I have described above fits into the framework of 360-degree feedback:

Most often, 360-degree feedback will include direct feedback from an employee’s subordinates, peers (colleagues), and supervisor(s), as well as a self-evaluation. *

This is not very common in science, but some leading institutes are trying it. For example, I have been involved in one the Institute for Cancer Research in London is offering to their faculty.

A 360-degree review paints a much fuller picture than the typical lists of Nature papers. And because most of it is based on interviews or questionnaires it wouldn’t even cost much money. I believe scientific institutions really miss out on a trick not offering such a type of feedback to their faculty. Every five years would be a good frequency, which in my institute would coincide with the quinquennial reviews we are having anyhow.

PIs can try to do some of it informally in their own groups. One of my colleagues had a session on their lab retreat called “The good, the bad, and the ugly” where group members had prepared anonymous statements of what works well in the lab and what could be improved. These statements were then randomly pulled from a hat and the PI sat on a hot seat commenting on them. Doing this every 2-3 years should give you a pretty good idea of what the critical issues in the lab are.

Another simple technique to get feedback is to have ‘exit talks‘ with everybody who leaves your group. Simply ask “If you could change one thing about this group, what would it be?” Because they are leaving anyway, they should be more open to giving you a frank and honest answer.

Increasing support and feedback

The issues I have flagged up here are not specific to science.

Some years ago Harvard Business Review ran a piece on Evaluating the CEO:

After I became a CEO, I was struck by how perfunctory the board was in its feedback on my performance. (…) My total worth was based on just three or four financial measures. *

Just substitute ‘financial measures’ by impact factor and grant income to see how similar the CEO’s situation is to a scientific PI’s.

Like PIs, CEOs tend to be strong-minded and independent personalities, who will do almost everything to defend their autonomy:

But to use CEO autonomy as an argument for limiting performance management to only financial measures makes little sense. *

Everybody needs support and feedback. And the higher you climb and the more independent you become, the less you will get it.

At the end of the day, my experience has taught me that it’s when CEOs think they no longer need real, credible feedback that they get into trouble. *

What I have described here are elements of a scientific leadership program to improve the levels of support and feedback PIs get. Some of it, every PI can implement in their own lab, but most of it needs explicit support by their department leaders.

So, one last time … Department heads, this is to you: Support your faculty!


7 thoughts on “Why science needs continuous leadership support

  1. Hi Florian,

    I enjoyed reading! A couple of thoughts:

    – I think the limited feedback for people in powerful leadership positions tends to reflect lack of courage on the side of their “subordinates”.

    – I think that the most powerful and effective way to improving the situation is creating incentives for group leaders to care more for their trainees. These incentives do exist but they tend to be much weaker than the incentives for getting top-tier papers and big grants. Unless we can change the rank order of the incentives, we will never change significantly the priorities and the behavior of group leaders / professors. Taking into account feedback from trainees can be a step in this direction but not without its own limitations and political overtones.


    1. Thanks, Nikolai.

      I agree, it can take a lot of courage to give honest feedback to the people you depend on in your career, unless they have invested in earning your trust.

      And yes, incentives are a problem. This is why I think the change has to be top-down. Institute heads (=the people with power) need to explicitly start evaluating their faculty by different standards than number of papers and grant income. And it has to be made clear to faculty early on what these standards are and that they include leadership and supervisor skills.

      The feedback from trainees can indeed be ‘political’ but it would only be one part of a 360 view. I think it is good if they contribute at all in some way.



  2. Pingback: Scientific B-sides

You gotta talk to me!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s