The Case for Lab Notebooks in CS Research

Posted on June 30, 2013 by ivotron

As part of their day-to-day work, many research practitioners from several disciplines (specially in the biological sciences) keep a lab notebook in which they record all their notes1 in chronological order. When the time to publish comes, they select the most relevant entries from their notebook and assemble the article/presentation/poster. This is a really useful practice that other theoretical (or "virtual") fields such as Computer Science don't generally follow and might benefit from it.

Personally, I've been keeping an electronic lab notebook as part of my PhD research and I've been advocating this practice among colleagues and friends in CS research. It's been really helpful not only for what I mention above but also in other aspects that I summarize in this post. In a subsequent post I'll share my workflow and the tools I use to keep my records.

Your brain works differently under different circumstances

Writing makes your brain work differently (matters of left vs. right hemisphere). Drawing too, so diagramming helps to understand concepts and explore ideas from different angles. I haven't used mind maps (I'm a keyboard-oriented person), so I haven't experienced the value of them yet. From what I know though, I think that a mix of outlining, note-writing and drawing might serve the same purpose.

The ends or the means?

tip 31 of 'Pragmatic Thinking and Learning':

Write on: documenting is more important than documentation (p. 190)

That is, even if the stuff we're writing won't be of any good/productive use, the fact that we write/draw things down makes us go through mental paths that otherwise we wouldn't take. I admit that at the beginning I was skeptic about it, being confident that the way I reason when I read a paper wouldn't be much different than when I write something down. It's not until I did it that I eventually found myself having 'Aha!' moments that otherwise would have taken longer to have (or not have at all).

Teaching is the best way to learn something

Arguably, teaching is the best way to learn: as you prepare to teach, you have to take into account the different point of views of the audience and anticipate questions coming from different angles. If you assume that the contents of your lab notebook won't be for your eyes only (even if they end up being so), you get some of the benefits of teaching a subject, i.e. by thinking not only in a single-minded way.

Do you remember what/when you didn't know? Neither do I

One of the common pitfalls in presenting complex ideas to others is to assume that the audience has the same background as we do, that they have read the same papers as we have and that a particular term is given the same meaning as we are. However, is very likely that only a few will be "in-line" with our thinking, so when we are preparing to present our work, it helps greatly to go back and see what things weren't clear when we began working on it.

I publish, ergo i do research

Last but not least, writing is one of the main ways in which scientists transmit their knowledge to others. We write papers and (for better or worse) are judged/evaluated for it. Thus, writing a notebook keeps our writing skills sharp.

The overhead

There's the issue of the overhead of maintaining a lab notebook. Instead of writing your latest entry on your notebook you could be reading the next paper on your to-read queue, analyzing the output of an experiment or implementing another piece of your experimental prototype.2 Personally, I think it's worth the price. Others might find it too costly. Also, as with any other skill, practice makes perfect.

On Openness

Notice that I'm not proposing to have an "Open" Lab Netbook. The main reason being that for many researchers (or researchers in-training as myself), the decision of what's public is out of their control (specially if you work with sensitive information, for a national lab, or something akin). In an ideal world, we would share openly without problems. In the real world, publishing your work might be detrimental, specially when it comes to finding "traditional" sources of funding.

  1. these notes include, but are not confined to, experiment methodology, preparation and results, as well as valuable comments on their findings. [My wife's] field is Molecular Microbiology and I've seen in countless occasions how many discussions among their peers are based on the content of each other's notebooks.

  2. By prototype I mean whatever form of prototyping one makes use of, be it throwaway or evolutionary.