I have the luxury of focusing on investigations’ intellectually exciting aspects, the planning, listening, conversing, tinkering, and problem solving because a couple dozen driven, highest caliber students and postdocs attack the hands-on. They undertake most of our group-to-group interactions. Having rotated through other labs and received varied training before joining up, they can consult counterparts elsewhere to solve practical problems as needed. My personal pace is too impatient for most collaborations. I rarely initiate them.

On the one hand, do not collaborate automatically. A young P.I. learns by venturing into the unknown to some degree, acquiring new abilities instead of immediately turning to colleagues. On the other hand, highly specialized experiments outside your area call for collaboration – but consider the big picture. Do not depend on collaboration for most of the project.

I collaborated very little at the start, far more so now. At the beginning, be extremely careful in associating your program with outside colleagues. If you must do it, approach persons from fields far enough away that it’s entirely clear who contributed what, e.g., an experimentalist might seek computational assistance. Be certain the other can be counted on for ultra-careful, accurate, timely work. This is difficult to determine from a distance.

When I was young I did straight chemistry projects and focused on independence. Funding preferences then changed. As a synthetic chemist, I had to collaborate with biologists to have access to any funding.

This is tricky. In my experience, collaboration very much depends on the culture of the department where you’re employed. Ask prospective colleagues their sense of this. I’ve collaborated extensively and productively throughout my career as encouraged by the institute where I am. In contrast, a colleague at a major university warns against collaboration, saying that the junior person frequently gets no credit for it. Results are attributed to the senior colleague only. Dreadful! Be warned. Incidentally, I’ve had a collaborator rip off an idea and give it to a competitor. It wasn’t the end of the world, but didn’t feel good.

I initially preferred not to collaborate although projects could have been made easier that way. If by chance given little or no credit for my idea, I reasoned, it would be difficult to change the impression this would make. Though hard when many demands are made of you, independence is essential. A few years down the road, after your expertise and research area are established, collaborative research is an exciting prospect that should be pursued.

Because my research is interdisciplinary, collaborations are very important. Work spread across several laboratories produces outcomes much more quickly than otherwise. That said, be very careful not to compromise your scientific standards and independence.

The theme for me from the start has been independence.

It is the scientific problem that must define whether collaboration is necessary, and with whom to collaborate. It is extremely important to choose collaborators carefully and well.

Before hoods flame up, a terrified examinee vomits, or your lab ceiling caves in, understand that essential collaborators include cleaners, custodians, supply room clerks, research librarians, child minders, and above all lab/departmental secretaries. These people will assist immediately, far beyond expectations, if given reason to like you. Routinely give courteous greetings, gratitude, and birthday flowers or six-packs as the case may be. Similarly your personal partner: however often non-scientists assure you they understand and respect crunch times, they still need preparing for pre-tenure’s duration and distractibility. All these collaborators multiply your capacity and influence. They introduce you to factors important elsewhere. Outsiders think it peculiar that colleagues often in competition for resources and recognition also write papers or proposals together, but that’s how it is.

As long as collaborations complement your expertise and maintain good communication and realistic expectations, they will be beneficial. Personal chemistry is another important component that should not be underestimated. In my own career, I start new collaborations openly and optimistically. Most often (although not always) they work out.

An effective researcher is not only skilled and self-aware in her own right, but also understands that a central element of collaboration is sustaining knowledge of others’ work. Keep a close eye on the forefront and networks of your field. Today, while it is essential that each researcher in a collaboration maintains visibility, partnership is equally integral to most programs.

At the beginning of your career, seek support, especially among foreign counterparts through bi-national BSF and GIF grants and the like, but do not collaborate with well-known P.I.s, especially those formerly your own. Once established, do as you wish.

Here too I take a combined approach. With collaborators you can stretch out to new areas without having to do everything yourself. Still, ensure your specific contribution and expertise are recognized as yours, not the collaborators’.

This question is difficult to answer as I rarely sought collaborations. For a long time, I did not want them. Nowadays, interdisciplinarity is increasingly important. To be part of such work, it is necessary to secure external expertise. I do think, though, that independent projects are essential to affirm yourself in a particular field of research.

Program reviewers weigh technical and intellectual contributions differently. For the most part you’re not expected to master all applicable skills; in fact, realizing the applicability of techniques you lack, locating someone who has them, and persuading him or her to assist are all to your credit. Conversely, core intellectual material must come from you.

In my field, pooled efforts with scientists having converging or complementary interests are crucial to achieving ambitious scientific goals. The same applies to teaching and mentoring broadly. All these depend on shared competence, mutual trust, and cooperation. I have had excellent collaborations.

The question is not whether to collaborate, but when and with whom. Naturally the answer depends on the project and prospective collaborator/s. Collaborations work best when all parties benefit and communication stays clear, constant, and honest.

I now collaborate extensively, and thoroughly enjoy it. At the beginning it would have been more challenging. I don’t discourage it as long as your contributions are completely distinct and paper authorship was well defined to start with, but, even so, recommend that most work by assistant professors be their own. Tenure is compromised when a significant percentage of the work is collaborative.

Unless a starting assistant professor is hired somewhere she can quickly muster a fairly large, diverse group containing at least one or two off-scale students, it will be difficult to solve an intricate, important problem fast enough to make an immediate difference to her new career. One solution is complementary expertise. Mine was to work at the bench myself through my assistant professorship. A lot changed later thanks to talented students and strong collaborators. I always credit these people on the very first slide. But the first ten years were rough.

I’ve always collaborated, ensuring coworkers’ expertise did not overlap with mine. We agreed up front on how to share credit and, for sustained collaborations, alternated first/senior author positions to keep things fair, but expectations and requirements have changed. My pathways are not as available today. After my post doc I spent two years building a lab and instruments. I published almost nothing. I was given time to build my career, for which I am eternally grateful. Today I see grade inflation start in elementary school and reach advanced training. Lacking papers in top journals by the end of your post doc, you may not get a second look.

I consider cooperation very important. It broadens our view of issues and encourages exploration of new methods. Cooperation is not the end of independence.

I’m mid-spectrum on shared projects but do review articles or proposals with an eye toward identifying possible collaborators. With regard to reviewing, my group shoulders a larger portion of that work today. Sometimes I wonder whether generating ceaseless conscientious, timely reviews disposes editors and funding sources in our favor, but doubt it.

Collaboration depends on the project.

Good collaborations take time to nurture, but fellowships and grants require letters of supportive participation sooner than you realize. Even before accepting a professorship, start researching and approaching colleagues (not those with whom you’ve studied) of benefits to be gained by collaborating. To be honest, I’ve been burned. After a collaborator pre-empted my work by presenting it prematurely and, incidentally, as his own, my only recourses were informal: to set the public record straight as possible by discussing the incident with my colleagues and, privately, with one of his; and to sever the collaboration. I won’t lie: this was a delicate, difficult situation for a young professor, but subsequent collaborations have more than compensated.

My group collaborates, and has extensively, since our start.