Planning for flexibility in a research project
The beginning of a research project is often very exciting: you have a cool research idea, and it is finally going to come to fruition, hopefully giving you some exciting answers. But a lot needs to happen between point A (coming up with a way to tackle a research question) to point B (getting the answer to that question).
Some people are tempted to plan every last detail of their project. I’ve seen people plan their whole PhDs and when exactly they will do which portion of the work. Needless to say, I haven’t heard of a single case where everything went according to plan.
Other people take the opposite approach, namely, they don’t plan anything because they know it’s very difficult to predict how a research project will go. But they run the risk of getting sidetracked, pursuing one issue, and then getting delayed with their project.
Here, I will describe how we can plan for flexibility when setting up a research project. We need flexibility as well as planning, and here is how to put them together.
Planning with broad strokes
Start by planning your project with broad strokes. What are the main chunks of work your project consists of? In my experience, most research projects have four main parts: setting up your experiment, collecting data, analyzing the data, and writing a paper.
Try to estimate how long each part will take, given its main components. If you have no idea how long something will take, ask your colleagues who work with similar equipment and/or on similar topics to get an estimate.
- Set up experiment: 2 months
- Program scripts
- Test scripts
- Collect pilot data
- Get feedback from colleagues
- Collect data: 3 months
- Schedule participants
- Collect data for all participants
- Back up data
- Do administrative tasks related to data collection
- Analyze data: 5 months
- Preprocess data
- Analyze data
- Analyze data some more
- (You can always analyze your data more…)
- Write paper: 2 months
- Write a draft of the paper
- Make beautiful figures
- Get feedback on paper
- Edit paper
- Repeat steps c-d as many times as necessary
- Submit paper (and hope for the best)
(Note: I am not including the review process here. That’s a whole other stage.)
Total: 12 months
Plan in extra time
When you’re making time estimates, do your best to base them on your previous experience and on that of your colleagues. Nevertheless, things will go wrong, or additional things will come up. Thus, when planning a project, you need to plan in extra time. The general recommendation is to plan an extra ¼ of the time you planned for a section of a project. Based on the estimates above, a more realistic estimation would look like this:
- Set up experiment: 1.25 * 2 months = 2.5 months
- Collect data: 1.25 * 3 months = 3.75 months
- Analyze data: 1.25 * 5 months = 6.25 months
- Write paper: 1.25 * 2 months = 2.5 months
Total: 15 months
You should also add any conferences you plan to attend and vacation time. Even if you don’t add them to your initial planning, you should take these things into account when assessing the progress of your project. Your project is not getting delayed if you go to a conference or on holiday; you’re just doing normal things that all scientists/people do. But once you’re back to work, continue using the planning outline to avoid losing track of where you are in your project.
The extra time you’ve added to your planning should take care of any small hiccups and extra things needed for your project. But if big issues occur, you may need to adjust your planning accordingly. The point is that you always know where you are with respect to your project (i.e., you can keep the big picture in mind) instead of losing perspective and getting completely consumed by one issue.
Things won’t go as planned
If there’s one thing you can be certain of, it is that things won’t go exactly as planned. The pretty plan you laid out for yourself won’t be fulfilled exactly, but that’s okay because the goal is not to follow it to the dot. Instead, it is supposed to act as a guideline and help you notice when you’ve gone off track. In this way, you can notice in time and make the necessary adjustments instead of waking up a year later and wondering where the time went.
Sometimes it will be difficult to tell whether to keep pursuing a certain issue or to move on. In that case, the general advice is: if you’ve spent an additional ½ of the time you originally allotted to this part, you need to move on. In the case of setting up the experiment, this would mean that if you’ve spent 3 months (instead of the initial 2 months planned) on this, you need to move on. If you choose to keep working on this, you need to have a good reason for doing so. Also, you should be doing so consciously and not simply because you’ve lost track of time.
An expert opinion can also be extremely useful in such a case. Often we, the individual researchers responsible for a project, are too close to the process itself to be able to tell when we’ve spent enough time on an issue. The opinion of a close outsider, such as a supervisor, can guide us in deciding when to move on. I have often experienced this with data analysis: I could have kept analyzing a rich dataset and gotten sidetracked because I believe that the analysis I was currently running was very important and interesting. But my supervisor can remind me when the answers we have are informative enough and when I need to move on to the next step.
Plan for flexibility
When setting up a research project, we know that we will need to adjust as we go along. But it is helpful to have an outline to guide us in the process. We don’t have to stick to the plan 100%, but it can help us to notice when something is taking too long and perhaps suggest that we need to make adjustments. And if we complete a part of the project early, then we exceed the expectations we’ve set for ourselves, which is always nice.