 |
| Photo by Dean Ricciardi on Unsplash |
Because writing is a staple of formal education since the earliest levels, it can seem like a waste of time to reflect on how to create written artifacts documenting your research process. But we are here to implore you to focus on this practice as you enter 2023. Recording research is a process that is distinct from all other writing you have done. It goes beyond those notes you scribbled during inorganic chemistry labs, requiring you to document the justification for your experimental designs, concerns you have about prior results, etc.
This blog post was cowritten with my colleagues Drs. Joshua Alper and Ulf Schiller. Our fingers are crossed that our lived experience can help you identify changes, big and small, in your writing process that can lead to fruitful research gains this year.
Post Contributor(s): Marian Kennedy,Ulf Schiller, and Joshua Alper
The goal of research is to add new knowledge and help your disciplinary community identify how “things work” (i.e., make quantitative, predictive models). It involves making observations, reflecting on what is known, developing questions that will help illuminate the unknown, etc. To get to this endpoint, you will need to have documented your research process from the nucleation of your ideas through the planning and execution of the experiments to your reflection on what your findings will mean to the community. This involves producing a sufficient record that would allow another researcher to replicate your entire process in your absence and determine if they come to the same conclusions.
Reflecting the importance of this process, funding agencies are increasingly requiring researchers to develop detailed data management plans and to publicly share data as well as publishing federally funded research. While the most obvious purpose is to benefit the larger scientific community (now and in the future!), the process also will directly benefit you. Your research records are your second brain and will help you answer questions about what you did and why – you can’t email yourself three months ago if you have a question about that! As you carefully collect and document your research, you will be helping your future self develop a thoughtful manuscript, write your thesis, prepare a research proposal that meets compliance, or establish your employer’s claim to “First to Invent” or discovery.
As research mentors, we expect that early-career researchers in our groups will see the importance of the documenting process for the greater scientific community or at least be concerned about their future selves. However, our words are not often enough. What seems to be a better motivator is looking at the actions taken generations of scientists in setting up their research studies. These retrospective analysis of how these scientists phrased their research questions/hypotheses, designed their experimental protocols, and ordered their steps really helps us think about the process we will follow. Sixteenth century physicist Tycho Brahe provides a useful example. He is today considered “the first competent mind in modern astronomy.” He also received an estate on the island of Hven from King Fredrick II of Denmark. Brahe used this estate to build Uraniborg, an astronomical research institute. The proof that he was a paragon of planning is that he established a papermill on the estate to ensure that he would have paper to provide material for recording his results and would never be waiting for paper to arrive when he was ready to record. Kepler later used the data he recorded on paper milled on his estate to develop the laws of planetary motion. Today, of course, we have a greater range of tools and less concern about emergency shortages because of digital files (voice, numerical data sets, etc.). But it is just as important to treat a laboratory notebook like a research diary: make it a habit to record entries regularly and consistently!
Documenting your work will feel like it takes a disproportionally large percentage of your allotted research time; we estimate about up to 75% of research time is used by researchers in documenting their work, reflecting on those artifacts (‘thinking’) or disseminating the research outcomes. Only 25% of our research time actually involves “action time”, such as running physical experiments in the lab or generating new computational code. Your time will need to be spent writing into your lab notebooks descriptive titles for each experiment you run, detailing your experimental plans, justifying the designs of those experiments, recording the experimental parameters, observations, and raw data, analyzing the results, drawing conclusions, and many other tasks.
We occasionally get pushback when we suggest to students that they need to understand how to document research when those students want to pursue computational research studies. They don’t understand the need to ‘document’ the development of code. But the same ideas apply to computation and modeling. In fact, we think one of the best outlines of how to keep a laboratory notebook comes from a computational biologist - Santiago Schnell. He eloquently outlined the framework for documentation in his paper “Ten Simple Rules for a Computational Biologist’s Laboratory Notebook” (PLOS Computational Biology, 2015). The rules he set forth are:
· Learn your institution’s or laboratory’s notebook policy
· Select the right medium for your lab notebook
· Make a habit of keeping the lab notebook in your desk
· Record all scientific activities in your lab notebook
· Every entry should be recorded with a date, subject, and protocol
· Keep a record of how every result was produced
· Use version control for models, algorithms, and computer code
· Keep a lab notebook that can serve as a legal record of your work
· Create a table of contents for your lab notebook
· Protect your lab notebook
After reading these rules, we wanted to emphasize a few small things.
Be organized: In our personal lives, we know that being organized can help us reduce stress, sleep better, be more productive, and maintain better relationships with others. These observations also apply to our work lives. For Dr. Kennedy’s research group, Ross Economy initiated the formalized insertion of “being organized” in each laboratory notebook when he was a graduate student in the group. He wrote up a “getting started guide” for new group members that encouraged them to think about the organization of their notebooks. Examples of organization include picking a standardized sample naming procedure and stick to it, using an index to allow others to find key information in your notebook, and providing context every time you enter reflections like writing the citation for a journal article next to your summary. Your lab colleagues and your future self will appreciate these small actions.
Written communication can be more than the written word: One of the things that we all have found helpful is varying the documentation in our notebooks to include symbols (arrows), tables, sketches, and drawings in addition to text. All these examples are written communication and can be used to share information with your future self (or other researchers) who will be scanning through your notebook. Sketches are particularly helpful and when we searched we found many examples of this technique from prior generations of researchers, including Alexander Graham Bell, while attaching photographs or screenshots are commonplace today.
Less isn’t more. In a lab notebook, more is more: One common shortcoming that we observe is when the laboratory notebook becomes a simple activity log (A was done and then B) but lacks the explanations that are crucial to follow your thought process. So be sure to include your reasoning and interpretation for any experiments and results. Include your thoughts about what the results mean, and what the next steps are. No one ever went back to an old lab book and thought, “if only I hadn’t taken such detailed notes.” In short, write everything down.
Don’t be afraid to revise your initial analysis after reflection: In a laboratory notebook, you will be analyzing generated data. For most of us, this takes multiple rounds as we identify the most appropriate methods. The documentation lets you later trace all your steps and insights, which can help immensely in crafting the results section for a manuscript, thesis, or dissertation.
Decrease your barriers to recording: As we close this post, we want to challenge you to remove any barriers you have to using your notebooks. The more you utilize a laboratory notebook during your research process, the more useful it becomes. For example, you need to remove the barrier of lacking time to run the experiment and record the experiment. Put time for recording on your calendar and plan it in. Some researchers who use paper—and plenty do—feel encumbered by larger notebooks and would rather leave on their desk rather than bringing it with them to all their research sites. In that case smaller laboratory notebooks will be easier to carry. Dr. Kennedy has found that she enjoys the process of writing much more with the right pens. She has made documentation a “more joyful” experience by removing all the free career fair pens littering her desk and replacing them all with Pilot gel pens (0.38). Having uniform ink color and pen size has made her how her notebook looks more, too.
Drs. Alper and Schiller have moved their research groups to electronic notebooks since these are more assessable in their working process. For organizing laboratory records electronically, a good folder structure and meaningful file names are essential. This will help you keep track of and find your digital notes when needed. Dr. Schiller recommends Tiago Forte’s PARA Method as a starting point for developing your own electronic records system. In the digital world, it is also common to provide a README file that includes the origins of data and/or description of programs. This example of a README file (template) is a good example and follows best practices as defined by a number of data editors at social science journals. Also, validated templates for commonly run experiments are essential – you should never duplicate an old experiment into a new one in an electronic lab notebook. This leads to polluting your new experimental record with protocol details, specific observations, and data that pertain to the previous experiment you copied. So, start every new electronic lab notebook entry with a empty template devoid of details from previous experiments.
Whether you use an electric lab notebook, or a paper one, a logical consequence of this post’s title (“If you don’t write it down, you may as well have not done it”) is that single most important piece of equipment in the lab, even if you are performing computations on your laptop or colliding high energy particles at a particle accelerator, is your lab notebook! Don’t do any research without it by your side!
Check in with your lead investigator to ensure compliance: We have found that there is a lot of variation in how notebooks are kept at academic institutions. The lead investigator in an academic lab (normally a faculty member) will typically set the type of laboratory notebook being used (physical or electronic) based on their preferences and the requirements of their funding agencies. For many academic groups, the standard is still a physical laboratory notebook, but industry has largely moved to electronic notebooks. Either way, this legal document should allow an external researcher to understand your process from idea nucleation to the publication of your outcomes in journal articles. Before changing a notebook style, talk to your research advisor to make sure that you are compliant with any funding source. At the same time, always remember that your future self will benefit most of all from a well-kept laboratory notebook.
About the Contributor(s): Dr. Schiller is currently an assistant professor in the Department of Materials Science and Engineering at Clemson University. His research group uses computer simulations to investigate complex fluids, soft matter, and interfacially dominated materials. Dr. Schiller received a National Science Foundation CAREER award to conduct multiscale simulations of nanofluid assembly for smart materials design. He recently also received an NSF EPSCoR RII Track-4 Fellowship and spent a summer conducting research at Lawrence Berkeley National Laboratory with one of his current graduate students. Students interested in a career in computational materials science can find more information on the Schiller Research Group website.
Dr. Joshua Alper is the scientific leader of a biophysics research lab in the Encoded Library Technology group at GSK. His contribution to this post was largely completed when he was an Assistant Professor at Clemson University with appointments in the Department of Physics and Astronomy, the Department of Biological Sciences, and as a faculty scholar in the Eukaryotic Pathogens Innovation Center. His research group performed fundamental biophysical research on the cytoskeleton, cellular motility and the unique molecular mechanisms of pathogenic parasites that was funded by both the NSF and NIH. He earned his PhD in Mechanical Engineering from Massachusetts Institute of Technology, and did postdocs at the Max Planck Institute for Molecular Cell Biology and Genetics in Dresden, Germany, and at Yale University. His contributions to this post were, in part, based on Joe Howard’s notes, which were inspired by Will Hancock’s notebooks.