Among physicists “manager” is not the most popular word, albeit that all group leaders are in fact managers. Why this is the case may be more than a blog post on itself, and I do not intend to write that post. Having quite some manager friends, for example in civil engineering or management consulting, I think there is a lot we can learn from them. One thing is the use of Gantt charts, milestones and go/no-go’s. Often people dislike these methods, a representative response is this post from 2008 on this blog.

To be honest with you, I love Gantt charts. These charts are a method of planning a project. It is a table that lists activities vertically and time horizontally. Using a bar you can indicate when this activity starts and ends. When one step is crucial for the next you connect the end of one bar to the start of another. So if one step is delayed, you can immediately see how this reflects in the end of the project. Moreover it allows you to set “go/no-go” points and milestones. Milestones are finished tasks, for example building a setup, obtaining a publishable result and submitting a paper. In the beginning of my PhD I feared one particular scenario: an “endless” struggle on a project with insufficient potential. Good planning forces you to think of the potential of a particular project, and the time justified to explore the potential. It is essential is to realize that things will never go according to plan, so you have to adjust you plan to the circumstances. In this post I would like to share how I tried to keep my PhD on track.

In the beginning of my PhD there was not a clear task, like a project to finish up from former students. Hence, I had to start designing a first experiment and order equipment. We quickly learned that we would need a cryostat, which has months of delivery time. Moreover, you need quite some time to define your requirements and find the right cryostat that suits these requirements. A Gantt chart then nicely shows you a block of more than half a year, in which you are mostly waiting, discussing and getting quotes: not really a fulltime job. So it shows you that you need to think of another activity that helps you in the course of your PhD.Given this situation my supervisor thought it would be wise to think of a start-up project to get familiar with the topic, and possibly obtain some results in the first year of my PhD. Basically, the plan was as follows:

1. Do a short project in the first year, simultaneously order equipment for the other project and design/make a first test sample.
2. In year two, we would do some exploring measurements to see if the main project is feasible, with a go/no-go decision to decide whether the project can deliver the required results for a thesis.
3. Finish of the main project in year three and four, or do an alternative project.

What I liked about making this plan is that you get a clear view of what your supervisor expects. It was comforting for me to hear that my supervisor expects to be able to make a good go/no-go decision after a year of experimenting on the main project. Some projects and/or supervisors might require a struggle of two or three years with a project before deciding that it is not feasible. It is also good to know what will happen if your project will be a failure, is there an alternative or do we need to think and explore one.

In the end, my PhD hardly went according to plan, as to be expected in science. Nonetheless the Gantt chart helped making decisions and setting priorities. So what happened? After the first year, the situation was as follows: we had the equipment and samples for the main project, and some results on the start-up project. As the start-up project was promising and we had some more ideas, we wondered if we should allow an extra half year to perform the extra experiments we had in mind. The decision of postponing the main project for half a year has to be taken with care. On one hand there is a nice opportunity that you would like to cease, on the other hand  you can not wait too long with the main project as that might get you into the situation that it has to deliver all the results for your thesis. In my opinion, this is naïve to expect from something completely new. In the end we decided to continue the start-up project, and it delivered three nice results.

After doing the main project for almost a year, we had the go/no-go decision for that project. The progress was promising, there were some issues that need to be solved first, but we observed multiple interesting effects that deserve studying. Given the multiple interesting observations, and the amount of feasible ideas for additional experiments, this should result in at least two more results. Hence, we decided to go for it. In making the go/no-go decision it really helped to know what the demands are: how many results should I get, and within what time frame. I was also very happy to have the start-up project finished already: on one hand I was able to apply the experience from those projects to my main project; on the other hand, the results of the main project seem insufficient to fill a full PhD thesis.

One point of critique to this way of managing a PhD is that it reduces risk. Which sounds strange, as reducing risk is typically seen as a good thing. However, for science it may not be helpful as you do not get into uncovered ground without taking risk. With the current trend of quantifying scientific progress by the number of papers and their impact two things may happen: people may start avoiding large risks, hindering large breakthroughs; or the opposite happens, large risks are taken at the expense of the career of PhD students. Probably it is best to manage the risk, by carefully combining high risk activities with low risk activities, to make sure that every PhD student can write a good thesis, while doing exciting and risky science.

Follow