Before you say anything, I know that when it comes to developing collaborations with their fellow scientists some PhD students don't even get a look-in. This is often simply because their supervisor has his or her own hand very firmly on the rudder. But if you are lucky and your boss feels you can handle it, you could find yourself making first contact with a colleague in another lab.
To the uninitiated, it may not be obvious why a PhD student would want to establish a collaboration. After all, aren't we supposed to be conducting our own unique piece of research, albeit under supervision? Look at it from the point of view of your external examiner. They want to see a candidate who will be able to contribute to their scientific community: someone who can network successfully with their colleagues abroad, for instance. Far from being chastised for collaborating, you will be demonstrating the very skills required for getting on in 21st century science. The tiny number of single-author research papers now published is testimony to the need to collaborate.
That's all very well, but what's in it for you? We all know that science can be very competitive, so you may be understandably reluctant to share your results or even let the other party know of your existence. But those you currently view as competitors can, under certain circumstances, become collaborators. The best example of this is where you have something they don't. Then you can start to bargain with them for access to what they have that you need, be it results, expertise, equipment, or whatever. There will be more on handling the politics of collaboration in the next article in this series.
Other potential collaborators may not even care less what you are up to, but they may still have something priceless to offer you: They may have done some of the hardest work of your PhD for you. Imagine you are faced with a seemingly insurmountable list of experiments (you may not have to stretch your imagination too far, I would guess!). Then you discover that someone from a large and wealthy group, who works in a related field, has just casually knocked off the most dreaded of your experiments. What took them 2 weeks might take you 6 months by the time you learn the technique from scratch. All you have to do is ask them nicely if you could have the result or end product. The perception amongst PhD students is 'there's no way they'll just give it to you'. Believe you me they may well do just that. It's how science works. For the chance of a joint authorship, or even an acknowledgement on your paper, your fellow scientists will often gladly tell you their results or send you their stuff. The trick is how to go about asking in a way that increases your chances of success, but more on that in the next article.
In my experience, there are two sorts of collaboration. The first is where you bump into someone who is either working in your field or who is doing something that you think may be useful to you. At a conference 'bumping into them' may literally mean that--something jumps out at you from their poster or talk. In this case you need to make a quick decision whether it's best to get talking face-to-face or wait to make first contact after the conference from the relative safety of your computer. As a general rule of thumb, I tend to jump right in with fellow PhD students and postdocs, but I think first before marching up to scary professors. Your advisor may also have 'issues' with senior scientists that you are not privy to, so speak to them first if you are in any doubt. As well as in person, you can also 'bump into' colleagues in the literature or via a Web search.
The second collaboration type is where you have identified a need to collaborate in advance of knowing who may be able to help you. This is harder to approach as the onus is on you to find the best person for the job. Without inside contacts and inside knowledge of your community (enter your advisor), this may come down to doing a bit of good old-fashioned detective work. The dear old Internet can be a lifesaver if your advisor can't help you.
In collaborations of either type, you will want to work with someone who will be open and fair with you. It's pointless throwing your results away in a one-way relationship. Collaboration is all about efficiency--reducing the time taken to produce valuable results. Translated, that could mean that you might get to do the really interesting experiments that you never dreamed you'd have time to do before the end of your PhD.
In the next article in this series we'll be looking at how to go about making first contact with your chosen one and how to handle the politics that ensue if they want to make it happen.