Most of us learn the art of writing papers on the job, often a painful process. In this four-part series, I’ll run you through my step-by-step approach to writing papers and, hopefully, help make the process of writing your first (or next) paper, a bit easier. As always, if you have any alternative advice or comments that you think might help others, we’d be glad to receive your comments.
The vast majority of us don’t actually start writing our papers by writing the introduction, research paper abstract, methods, results, or any other section, but rather by creating a series of figures that illustrate the story that we are going to tell. With this in mind, my first piece of paper-writing advice to any student is this: learn to make a complete, compelling, professional-looking figure.
Plan Each Experiment To Be a Figure
When you are at the bench, you can end up performing experiments just to answer a particular question for your satisfaction. This can lead to you relying on the controls you performed in whole series of experiments you did prior to this one, and so you don’t feel as compelled to add those controls to this experiment. After all, you have been working with this reaction for months and you know that nothing happens to the substrate without Protein X. To compound matters, you are likely presenting a ton of experiments in your lab meetings, so your labmates and mentor may not even call you out on not including those so-called ‘trivial controls.’ You (or worse a reviewer) only notice when you first decide to communicate your story to people outside of your lab—whether that communication comes in the form of a talk, poster, or a paper. The lack of complete controls leads you to scramble and look through your lab notebook at every experiment that you have recently performed—only to realize that you don’t have one experiment that has every element required to satisfy a critical reviewer.
“Think in figures” is my advice to keep young scientists from falling into this trap. Instead of heading straight to the bench with your burning question, stop and ask yourself what elements (positive controls, negative controls, gel markers, etc.) need to be in this experiment for it to be a figure in a paper. Imagine that your hypothesis is correct, and you’re going to use the results of this experiment to convince a critical reviewer—what lanes need to be on the gel to make your point? Would you convince (or at least frustrate) an opponent of your hypothesis?
Not only can thinking in figures help prevent you from having to re-do all of your experiments right at the moment that you should be focused on other matters (i.e., writing your paper), but it will tighten up your science. The first time your science should be critically reviewed is before you show it to anyone else.
Become a Master of Your Techniques
By ‘techniques’ I mean the manner in which you visualize your experiments: agarose gel electrophoresis, SDS-PAGE, western blotting, immunofluorescence,and so on. Most of us start out with pictures that fall into the ‘uninterruptible’ end of the execution scale, and through advice and experience quickly advance to ‘ugly, but usable.’ Some people stop trying there and shift their entire attention to developing the story at hand. Some don’t believe that pretty data is worth the effort, while others think that generating pretty pictures is something they’ll do later, when they’re ready to publish.
For those who believe pretty data doesn’t matter, crack open an issue of Cell and compare the figures to those from a low impact-factor journal in your field. Obviously many factors determine where a paper will be published, however, I believe that ‘prettiness’ of the data can have a very real effect. Clean, sharp gels and high signal-to-noise ratios in antibody-dependent techniques inspire confidence in the reviewers, which influences where the paper will be published. So-called ‘pretty’ images also sways the readers and affects the actual impact of your paper in the field. As much as we might hate to admit it, a large part of science is trust in the authors of the papers we read, and you want people to trust the results you generate.
Becoming a master of a technique can take as long, and be as frustrating, as developing the story itself. Therefore, I always advise a student to strive for this goal from the very beginning. Now, I’m not talking about taking a month off from project-advancing experiments to perfect your electrophoresis skills. You will have plenty of chances to practice while advancing your project because even generating the first real figure of your story (paper) could take an entire laboratory notebook’s worth of work. After collecting the data from any particular experiment, analyze the execution of the technique as well as the scientific data generated by the experiment. If you think that the data could be ‘prettier,’ then think about what you might be able to do to improve your technique when you plan the next experiment. You can shorten this learning curve by keeping an eye out for other scientists inside or outside of your lab that already have mad skills at the technique in question and ask them for advice and help.
Learn to Use a Vector-Based Graphic Design Program
We’ve all heard the cliché “a picture is worth a thousand words”. In science, often a picture is irreplaceable with words, and a good illustration is worth 200-800 words. (Yes, I actually tried quantifying it a couple of times.) With the restrictive word counts of some journals, a handful of good illustrations can buy you several more paragraphs of analysis and pontification in your manuscript.
Many scientists use PowerPoint to make their illustrations and, depending on your project, you can get by with it. Considering that it is a presentation program and not a graphic design program, it does a pretty good job. But I have never met anybody that regretted the time that they spent learning to use a true vector-based graphic design program like Adobe Illustrator or CorelDRAW, among others . These programs offer more options and control. However, these options come with a dizzying array of buttons and commands that can be intimidating.
I would advise any young scientist (or even a not-so-young scientist) to spend the time to become proficient with one of these programs. Some of the features in these programs are intuitive, but there will be a significant learning curve for the more advanced features for most of us. Therefore, you should learn to use this program when you have the time and not when the pressure is on to publish your paper.
The good news is that there are several options for learning these programs. The most basic is the ‘Google method,’ where you plug away at it, and search the internet for information about features you encounter or methods for accomplishing the task you have at hand.
A better option would be to buy a tutorial book that shows you the features by stepping you through a set of lessons.
The best option is to take a class. If you are a graduate student, you may be able to take undergraduate courses at your university for free. Alternatively, community colleges often offer courses on popular programs at very reasonably prices, although they aren’t free, and some are offered at night or on the weekends.
Once you have the basics down, if you learn of others that are proficient with your program, don’t be shy about asking for help—I can often show somebody the solution to their problem in minutes after they have spent hours (and hours!) searching for the answer.
Generating compelling, polished figures is the first step in communicating your results to the world. They will not only become the basis of your paper, but will be the starting point for any posters and talks that you give on this project.
In the next article, I will discuss the next step: writing the first draft.