How do you do college notes with Zettelkasten — meta post

Hey, I’d like to hear about how you’re using Obsidian and Zettelkasten (ZK) to take college notes. As far as I’m concerned, there’s two main ways to go about this:

  1. You have all the notes in ZK+Obsidian. You prepare for exams from your ZK notes — all the details are in there.

  2. You find (1) too time-intensive. You take notes in a linear fashion (on paper, on iPad, or whatever) and bother organising most of them into ZK. Your ZK has only the most important insights and a-ha moments.

Do you do (or would you recommend doing) 1. or 2.? What’s your workflow? I’d be happy to hear anything related to the topic of taking notes in college!

I’ve been discussing with @nickmilo about what’s the best way to do 1. — that is, what’s the ideal system under which to take college notes in ZK. But I think it might be worth to discuss also whether it’s worth to apply the system to everyting.

3 Likes

I’ll go with 1. even if I don’t have that much time.

I basically make a Zettel for each concept, but I leave most of the notes unfinished (for example the proofs of the theorems) because I can consult the notes later.

I have:

  • A lot of example notes: A main example with links to all the properties that it satisfies.
  • Theorem notes: I have the statement, aplications (with examples) and a link to the proof (or the proof it is short). I try to be very descriptive in the title.
  • Definitions.
  • TOC: I start with the teacher’s outline, but later I remove some unnecessary things as they can be found following related concepts.

I often find a lot of relations that they are not explicit in the notes. The main ones are:

  • Examples that are used in different contexts.
  • Theorems that are related to other ones (generalizations, particular cases, …).
2 Likes

Good topic. College studies are arguably the best time to start “taking smart notes.” I certainly wish I had these ideas when I started my undergraduate!

It may be worth looking at other resources on how to study. I found Barbara Oakley’s A Mind For Numbers a fantastic read, and I only read it just before writing my PhD comprehensive exams a couple of years ago. I’m sure there’s others out there.

Oakley’s method is simple: rewrite ideas in your own way, work through the logic of the concepts step by step and ensure you understand it completely, take a break when you’re stuck (+ study early and constantly so you can take breaks), and genuinely quiz yourself so you can assess your gaps in knowledge. (Her writeup is much more useful than mine, though.)

I really valued the creation and study of good flash cards in my PhD. If I redid my education, one thing I’d do is funnel all of my notes into something like Anki and study with spaced repetition. I imagine there’s a way to “feed” notes into Anki, so that might be something worth exploring.

Otherwise, I’d be trying to keep a forward-looking repository of notes that might be useful later on. I have pages and pages of OneNote notes from my psychology and computer science degree, and thousands of highlights from my master’s of design, and none of those are in a useful form. Having started reading Ahrens’ Taking Smart Notes, I wish I had captured the most interesting ideas in my own words (with reference info) when I’d first come across them, and built up maps of content to knit these ideas together over time. I don’t think this would have take more time than I used to study anyway, but it would have made it more useful in the long term.

11 Likes

I graduated with my undergrad in accounting in January of 2018, and then ended up passing the last part of the CPA in September of 2019. All of my note taking was done on paper, with physical flash cards sometimes used as a means to test myself. I was pretty against typing notes throughout all of college, and I still do feel that typing notes is less effective. The issue that I have come across now is that I do not remember the vast majority of the topics I covered in college.

This is why I’ve pivoted to taking digital notes.

I’m still changing how I do things every few days, but I definitely fall under the first group. All of my notes are within Obsidian, and I create zettels as I read through the textbook. An exception to that are example problems which are on paper. I mark each page with a number, and include that number in the zettel on that topic. I’ve experimented with highlighting first, and then making notes, and I’ve also created more traditional notes and then broke them down into zettels. The issue with these two methods are that they consume a lot more of my time, and don’t really seem to provide any sort of benefit. As far as I’ve noticed, the main benefit from this entire system is the step of rewriting the notes in your words. You only really see that take place in the step of creating the zettels.

I do think that note taking on it’s own doesn’t do much for long term memory. Something that’s been repeated a million times and something I can confirm is that testing yourself after the fact is key to remembering things long term. While this method does encourage you to reference back to your materials, it probably doesn’t do enough.

That brings me to the final goal which is to incorporate Anki in such as a way that almost every zettel I create will have an accompanying flash card.

My classes start up tomorrow, so I’m sure I will touch base again in the future with how it goes.

4 Likes

Thanks for the answer! I’m glad you connected the note-taking part of the question with the learning part.

Good topic. College studies are arguably the best time to start “taking smart notes.” I certainly wish I had these ideas when I started my undergraduate!

It’s always too late! I’d be grateful to have these ideas before I even joined the college; now I’ve “lost” two years and must make up for them retroactively. However, it’s the college that inspired me to look for better ways to learn and take notes — I suddenly realised that even from classes I straight-A’d I don’t remember much of the stuff, and that my notes aren’t very useful for relearning it.

Speaking of college, I sincerely believe that stuff like this should be taught; there should be a two/three week class on learning methodologies and note taking.

It may be worth looking at other resources on how to study. I found Barbara Oakley’s A Mind For Numbers a fantastic read, and I only read it just before writing my PhD comprehensive exams a couple of years ago. I’m sure there’s others out there.

I didn’t read the book, but I went through the famous course Learning how to learn (from Oakley as well). I learned much, but I must confess I fail to apply most of it in practice.

I really valued the creation and study of good flash cards in my PhD. If I redid my education, one thing I’d do is funnel all of my notes into something like Anki and study with spaced repetition.

Even without flashcards, I barely manage to keep up with the coursework now. Do you think it feasible to make flashcards for all the classes, in such quality that you can then use them for exam preparation?

I like the idea of flashcards (they’re like a hack for our brains), but I don’t like the disconnectedness of them; I’d still need to have some underlying interlinked material that I could read through to understand the facts I’d learn from flashcards.

Given that writing good notes is very time consuming in and of itself, and so is making flashcards, I opted for “notes” instead of “notes + cards”. I’d love to hear a different perspective on this matter, though.

The time dimension isn’t all there’s to it, however. I think that flashcards are great for stuff that you’d like to remember, long-term. There’s one more way to remember something long-term: using it (either applying the knowledge, or learning something new that build upon the old thing).

So, I chose the strategy of writing good notes on everything and preparing for the exams, after which the knowledge falls into two categories:

  1. Knowledge I need for my job or for further studies. This will automatically stay in my head, because I’m using it.

  2. Knowledge that is good to have, but not critical to anything I’m doing at the moment. Once I need to use it, I can get a refresher from my notes, provided they’re good.

I think for PhD flashcards are more appropriate, because you need to have a lot of knowledge in the working memory at once — and you know you’ll need it all. In bachelors (which is where I’m at, now) there’s a wide spectrum of classes, most of which are only remotely related to what I’m doing at my job at the moment. Nevertheless, I’d love to hear more about your flashcard framework and about the ways you think it could help me.

[speaking of past notes] none of those are in a useful form.

Do you mind sharing why? Maybe there’s a lesson I could learn, some problems I could dodge.

@Eloitor Thanks for the reply. How do you name the files? Until Obsidian disconnects the note titles from their filenames, I have to formulate the theorem’s (or proof’s) title in such a way that describes it contents — and that’s often hard.

@phil Thanks for the insightful answer! Have you seen RemNote? I’ve never used it myself, but I think it’s similar to Roam with an added deep flashcards integration.

I do think that note taking on it’s own doesn’t do much for long term memory

I agree, but I also think we don’t need as many things in our long-term memory as one might think. Please see my reply to @ryanjamurphy above; I find writing good notes time-consuming enough, and I can’t imagine making flashcards for every fact in every class on top of it.

I should consolidate my approach to flashcards into one blog post. As yet I can only point you to my discussions on the Mac Power Users forum:


You will probably be jealous of the interlinkedness afforded by DEVONthink and Anki together. It should be possible to use a similar linking system to tie in Obsidian notes, although DEVONthink’s universal unique identifiers and URL scheme made this process really nice.

In general, though, the most value you generate from flash cards is garnered from creating good ones. Taking good notes should get you 80% of the way there. Processing those notes over time will get you another 10-15%, because you’re effectively reviewing with spaced repetition. The only thing missing is self-quizzing, which you could just do in your head.

Do you mind sharing why? Maybe there’s a lesson I could learn, some problems I could dodge.

My notes from my B.Sc. are effectively just transcripts of what instructors said. I might as well read modern textbooks. If I did it again, I’d also try to collect the most interesting concepts, the biggest questions I had, and other forward-looking classes of notes in some fashion.

My notes from my M.Des. are in highlights and comments on hundreds of PDFs. I never extracted nor processed or organized them into a more useful form.

I find writing good notes time-consuming enough, and I can’t imagine making flashcards for every fact in every class on top of it.

I would try to find a way to build flashcard creation into existing workflows. The flashcard workflow I describe over on MPU involved programmatically generating flashcards from PDF highlights. Arguably a similar concept is possible via Obsidian notes! That said, I wouldn’t necessarily prioritize this (see my comment above in this post about creating good notes getting you 80% of the way there).

1 Like

In general, though, the most value you generate from flash cards is garnered from creating good ones. Taking good notes should get you 80% of the way there. Processing those notes over time will get you another 10-15%, because you’re effectively reviewing with spaced repetition. The only thing missing is self-quizzing, which you could just do in your head.

That’s what I was suspecting, thanks for putting it in (better) words. I’ll definitely look up the DEVONthink+Anki integration, but I won’t stress it. Thanks for introducing me to those forums, they shall come in handy during some future exam periods, when I’ll have solved the Zettelkasten problems ;–)

My notes from my B.Sc. are effectively just transcripts of what instructors said. I might as well read modern textbooks.

Good point, I’ll be wary of that. To be honest, the things the instructors say can be themselves so often perfectly substituted by textbooks that I wonder whether there’s any point being at the lecture at all… Just some thoughts.

1 Like

I name the important theorems as in the literature, but if there is some collision I specify further. The file contains the statement and consequences / examples. I write the proofs in a separate file. (I append “proof” to the name)

  • “Shannon Theorem (Noiseless coding)”
  • “Sardinas-Patterson Theorem”

For theorems that are not very famous I write a description. (Here the proofs are usually contained in the same file)

  • “the entropy lowerbounds the average word length for uniquely decodable codes”.
  • “minimum distance = minimum weight for linear codes”
  • “dual of a linear MDS code is MDS”
  • “dual of a cyclic code” (sometimes I write what it is expected from the theorem but not the result itself).
  • “upper bound for the entropy function”

To name objects I use plurals when there are more of one object of its kind and singular if there is only one. Here I write definitions, properties and examples.

  • “linear (block) codes”
  • “Reed-Solomon code” (There is one for each finite field)

I think it would be useful to have an example of complete vault of a subject.

1 Like

I have seen RemNote… but I think it needs a bit more polish before I can really want to use it. The way they handle flashcards is interesting, but I didn’t enjoy using it as much as I wished I did!

As for making flashcards… it can be time consuming. What I plan to do is create the cards as I make notes. I’ll be creating the cards in Excel and then importing them into Anki.

There are apps that can make integrate flashcards into your workflow easier. Polar Bookshelf can make an Anki flashcard directly from a highlight within the reader. That can easily save you a lot of time.

Seeing a lot about flashcards and I wanted to point to this thread here: Simple spaced repetition plugin in which there is a discussion about a video on using neuracache for spaced repetition (the method utilized by remnote) to create flashcards. I think it’s worth taking a look into.

I checked out this thread because I’m an engineering student and considering all the problem sets I handwrite and all the textbook information I never get to, I was really interested in hearing other’s thoughts on college note-taking.

Also sorry if this comment is a bit messy! This is my first comment ever and I’m still new to obsidian.