With the first midterm graded and returned, many of you have asked for input about how to improve going forward. We're at the halfway point in the quarter, so in the grand scheme of things the night is young: there’s plenty of time to solidify your skills, and even if you had a rocky start to the quarter you can end strong and be happy with the grade you receive.
The unifying theme of this handout is that you can get dramatically better outcomes from CS103 – or pretty much any class – if you’re strategic and intentional with how you use your time.
Much of the advice in this handout is of the form “consider spending thirty minutes a day doing X, Y, and Z.” I know that most of you are already putting a ton of time into this course and are making a good-faith effort to master the material. I don’t want the takeaway from this handout to be “you need to put even more time into CS103 if you want to succeed.” Quite the contrary. If you set aside some time every day to focus on high-leverage areas, I think you’ll find that you spend much less time overall in CS103, since each unit of time you’re spending is significantly more productive. The question is how to spend your time more productively. Here are a few concrete suggestions.
Revisit your prior work and actively patch up any holes in your understanding.
In CS103, it’s very easy to structure all of your learning around trying to get through the current week’s problem set. While this is a useful way to learn new concepts, if you’re not intentional with how you approach the problem sets you might not end up getting enough practice with the skills that you’re weakest on.
When you get back a graded problem set, pick a problem or two that you didn’t do as well on as you would have liked. Read over the grader’s feedback, review your work, and then take thirty minutes to rewriting the proof to address that feedback. Take what you just wrote and either stop by office hours or post it on EdStem to get the TAs to look over it. Specifically mention what issue you had on the problem set and see what the course staff thinks about your new version. If they think what you have is in good shape, great! You’ve internalized the feedback. If not, take this new piece of feedback and try rewriting things again.
This process will take some time – I figure you’ll spend about half an hour a day working through this – but it’s one of the single best things you can do to improve your performance. And from experience, this will dramatically decrease the amount of time you spend on the problem sets. Chances are that in fixing things you’ll find that you can tune the way that you approach problems in the first place, which will focus your efforts more effectively, or you’ll find that you can write half as much as you used to, increasing your overall throughput.
Proactively ask questions about material from lectures.
It is extremely easy to get into a trap in CS103 where you fall behind on watching lectures, which means that you have to scramble to fit them in before the next problem set, which means that you fall behind on the next week’s lectures, compounding the problem.
If you aren’t regularly attending lectures (or watching them remotely on a fixed schedule), I would strongly recommend trying to get into the habit of doing so. We know this takes a lot of time, but it is absolutely worth it. Once you get into a good rhythm of staying on top of things, you’ll have more time to pore over the material you’re learning and think about how everything fits together.
If you are staying on top of lecture, I recommend spending thirty minutes a day actively reviewing what was covered. Here’s one useful way to spend your time: pick a proof we did in a lecture. Look back at your notes to see what the key insight or insights were. Then, without looking at the written proof from lecture, take a stab at rewriting that proof completely from scratch. If you’re having trouble doing so, that probably indicates that there’s a concept that didn’t click. Great! Review your notes again, pull up the slides, and if you still can’t crack it, post a question on EdStem with what you’ve tried.
Manage to get a proof working? Great! Compare what you have to what we wrote in lecture. Does it look similar? Is it totally different? Your goal isn’t to match what we have word-for-word – that’s pretty much never going to happen – but rather to see if you can reconstruct things based on your understanding of the ideas. If what you have looks really different than what we have, post it on EdStem or stop by office hours to ask for input. Maybe what you have is equally valid and just follows a different line of reasoning, in which case, fantastic! You just learned a new approach to understanding the theorem. Or maybe what you have doesn’t work because of some minor nuance. In that case, even better! You just learned something new to watch out for, and probably got some feedback on your proofwriting approaches in the process.
Keep the TAs in the loop when studying.
We release a ton of practice problems around each exam. For example, for the second midterm, we’ll release Extra Practice Problems 2 and will give out another set of practice midterms.
As you’re working through the practice problems, we recommend getting the course staff to help check your work. If you’ve worked through a problem, feel free to stop by office hours to get the TAs to review what you have, and take their input seriously! If you’re getting feedback of the form “that’s the right idea, but be careful about how you’re writing it up,” take another stab at writing up the answer and see what they have to say. If you’re getting feedback of the form “there’s a logic error here,” ask the TAs how they arrived at that conclusion, make a note of their technique, and see if you can start applying that technique to your own proofs in the future.
Work in pairs, but do so strategically.
We encourage you to work in pairs on the problem sets because if you’re deliberate about the way you work in a group, you can learn much more than just flying solo. For example, if you write up a draft of a proof for a particular problem and hand it to your partner for review, you can get their honest feedback about what they understood and what they didn’t, which might help you tighten up your presentation or expose subtle logic errors.
If you haven’t been working with a partner, we strongly recommend partnering up for the remaining problem sets. If you’re not sure who to ask, post on the pinned partners thread on EdStem.
If you have been working with a partner, take a step back and reflect on your team dynamics. Are each of you doing all the problems independently (good), or is your partner always taking the lead (not so good)? Are you jointly working on the writeup (good), or is all the wording your partner’s (not so good)? Can you use your partner as a sounding board (good), or do they always give away the answers to the questions (not so good)? It can be easy in any relationship to arrive at a dynamic purely accidentally, but by being intentional about how you want your interactions to go, you can end up with significantly better outcomes.
Need some more targeted advice?
We'll be setting aside some time this week for 1:1 meetings, so if you'd like someone to debrief the midterm with you and offer some additional suggestions for what to work on in the next five weeks, send Vyoma an email at vyoma@stanford.edu and we'd be happy to chat.
Best of luck going forward!