How to take notes from scientific literature
A bit over a week ago, I took my licentiate examination. One important step in earning a PhD from a Finnish university. This exam has been scaring me since the very first day I started my PhD, like a lurking shadow in the back of my head. The licentiate is a pre-diploma to a PhD doctorate that shows one has a strong and deep understanding of his/her own topic. Also, it shows that one is capable of placing his/her own topic into a larger context. They recently changed the format to a mid-term presentation of your progress, with opponent evaluation–a sort of mini PhD defense. However, I had to take the old exam: a 7-days long literature review with writing of opinion pieces on various topics. For myself, that would mean reading a lot of scientific literature in a short amount of time, understanding it, digesting it, and the cherry on top, writing in English! So far all of my work has been read and edited by my supervisors. They usually had quite a bit of work to improve my English. Scary.
How taking good notes saved your precious time
My examination consisted in reading 14 papers that counted for about 260 pages and writing three essays of about three pages each. I knew the biggest challenge could actually not be the writing itself but reading that much scientific literature in less than 7 days. An average reader reads popular novels at a pace of 200-250 words per min. I read scientific literature at a pace of 30 WPM! Yes, I am a sea turtle when it comes to reading scientific literature. It takes me time to understand the ideas, the methods, the stats, not to mention each authors has a different style of writing, some have bad styles. Usually, I would have read the literature entirely before to start writing to get a good overview of the different topics. This time, I went for a different strategy. I took the bull by the horns and the problem head on: I read all the abstracts, picked three articles that seemed relevant for all the questions I had to answer and read only those before to start writing already on the second day. That way I would stress less when I would run out of time and would need to write. I repeated the process: reading, writing, shaping, reading, writing, shaping.
What really helped was to decide on a gross skeleton for my essays just after discovering the topics. I brain dumped all my ideas on paper for 1h and was refining on the go as I saw what was missing, where I could go more in depth and what could be cut out or transferred from one essay to an other. The literature I had to read seemed to be particularly picked to fit in the summary of my PhD so I took notes on the way. Pages of notes. Obviously, it slowed me down at first. Halfway the time, I was not halfway the literature. But I knew it was gonna pay off somehow. When I reached the time I should really get going with the writing–around Friday–I could just pick ideas from my notes and wrap things around those ideas. I wrote 60-75% of my essays on Friday and Saturday, leaving Sunday for improving the writing, editing and piecing things together. Now that I have those notes, I know that I can use them later in writing scientific articles if I need.
Going manuscript or digital?
So more concretely, how do I take my notes. I’ve tried many things over the years. At first, I would print the most important papers and start highlighting everything that seemed important. I usually would end up highlighting 80% of the text when the main idea is probably contained in less than 40% of the text. On top of being not so environmentally friendly, it would not be so efficient as I would not recall the main ideas after months when I would need the papers. If I would be lucky, I would remember that I would need this or that paper for this or that ideas but would need to read the paper again (or at least 80% of it) to find the ideas I’m interested in. A long and painful process that would mean keeping a pile of 30-40 papers next to me during the writing. One’s desk can already get quite messy when one’s want to stay on top of the literature so imagine keeping piles of papers on a desk for months. I then tried a different strategy: writing the main ideas down on a piece of paper, have a codification in my literature manager to be able to retrieve those notes and keep all those notes in a clipboard. When I’d need them, I could take out all the different sheets I need, put them in the order of the skeleton of my article, and write from there. If it would mean retrieving more easily ideas from different studies, I quickly started piling up such sheets–usually one per main ideas so several per papers–that it would take a lot of time to organize them into big topics, even more when topics would overlap. On top of that, I would not be able to search within those notes easily nor be able to have them on the go. Not practical to carry around clipboards full of notes.
From there, I decided to go for a digital equivalent. I used Google Keep as my note taker and would photograph schematics and graphs if I’d need. Despite loving Google Keep, its user friendly interface, and connectivity with other Google apps–I use it on a daily basis together with the Google Calendar–it is not really designed for keeping large notes, more for short and short-term notes like shopping lists or To-Do lists. On top of that, it is impossible to change the font settings to make your notes look good or to highlight things. I found plugins to inject markdown into Keep notes but it was dodgy and more of a hassle to use it in combination with Keep. As I was taking notes for my Licentiate examination, I gave a try to Evernote. The app comes in free and premium subscription-based versions. I was afraid to be limited in the free version but so far, I really love it! On the long run, I might even subscribe to their services. Compare to Keep, you can totally customize the text from font colour, font size, have in-line images so that one could add shots of drawings and graphs. One of the cool features is the ability to make tables or even add pieces of codes. Perfect for keeping bits of code from R statistical program. There is so much I could tell about this way of taking notes.
What your notes should contain
- A reference code to the note itself (to paste for instance in your reference manager)
- The full reference in whatever format you prefer (at least authors name, year, journal abbreviation to be able to retrieve the source)
- The aims of the study (translated in your own words, or copy-pasted from the aim section of the article)
- Your notes themselves (main findings, methods…)
The take home message
- Start to take notes as early as possible
- Do not just highlight the article, take active notes
- Best is to go digital or to take both digital and hand-written notes
- Stay consistent with a system: a template with reference code, useful info