Comment by abdullahkhalids | Hacker News Reader

Comment by abdullahkhalids | Hacker News Reader

Parent Full thread

abdullahkhalids

I can give you prof perspective on this. Going between different stages of education, you can't really double or triple the amount of time you study, but you need to change your study and thinking techniques to learn 2x or 3x more [1].

The problem is that if the teachers try to set up too easy a ramp at the start of a new stage, many students refuse to upgrade their study/thinking techniques. They continue to use suboptimal methods from their last stage, because its easier. But before they know it, they progress to a point where their suboptimal techniques are wildy unrealistic. At this point, they try to upgrade, except its much harder to upgrade your techniques at MATH301 then it is at MATH101.

One way out of it is to jolt incoming students with assessments where the old techniques simply don't work, and they are forced to upgrade immediately. It is kind of cruel, but it works well.

[1] For instance, a freshman PHY101 Mechanics course covers the same material as 1-1.5 years of high school with higher difficulty.

Could you elaborate on what those "upgrades" look like?

At least for math, I found that by around Calc II I had to start grinding problem sets to do well on exams, and I can't imagine what the next "level" would look like for something like a graduate-level math course.

There is no concise general answer to that question, especially one that covers many fields.

What you are saying is correct - from grade 1 to phd, the best way to learn is to solve problem (sciences) or create art pieces (arts) etc. As you solve problems, there are a few things I can say about how you can upgrade.

* If you are level N, you should only be learning techniques at level N. If you find that you are struggling even a little bit with level N-1 techniques, adopt an immediate no-nonsense attitude about eliminating those confusions. Eg. I have seen far too many students in my second year(!) differential equation struggling with solving with quadratic equations. This means they have to constantly jump between different levels of abstraction (algebra and differential equations) and that makes the question much harder to get right. A few hours of serious review should eliminate any confusion for something that probably took 2 weeks in high school, and probably give them an entire grade bump in differential equations.

* Figure out the meta-techniques at level N and take-off the training wheels. For instance, in graduate level applied math, you are working with a lot of theorems. Something that needs to become part of your study is generating positive and negative examples of each theorem as soon as you encounter it (something the book/prof did for you in undergrad). Nobody is ever going to teach you this at the grad level, as there is very little a grad prof can say to help you learn such meta-skills. But they will hit you with novel theorems in exams (or you will encounter new theorems in research), and you need to have the skill ready.

* Length of each study session and intensity of your study. In high school, a smart student can watch TV and still learn everything for an exam. A undergrad can listen to engaging music while still thinking about their problem on an assignment. A grad student needs the discipline to sit in a quiet room and fully engage with the problem for several hours. A PhD student might need to think about the problem and only the problem from the second they wake up all the way to when they go to sleep. Andrew Wiles might have to lock himself away for 6 years and immerse his whole life and being into solving Fermat's last theorem.

P.S. Not very happy with my answer. I need to chew on this for a few days or weeks.

Jolting people with assessments where old techniques don't work is fine as long as you don't do it first time on exam.

But also, students here just want to know whether they are expected to give detailed reasoning or less detailed one. How many questions they need to answer to get A. Basic stuff about rules.

A lot of courses can indeed be improved by being more legible in expectations. But there is actually a counterpoint. As you make your course more legible past a certain point, there arise the "lawyer" students. They will use the rules laid out to weasel out of work, to to get a grade bump - long and tedious bad faith argumentation.

Its actually the kind of thing you don't see till you actually become a teacher and try to make these sort of changes. The only way out is to have a a minimal amount of illegibility that allows the prof to actually stop these students in their tracks. Time and again I have seen new profs try to be legible and then realize a year later that there is a limit to it.

In general, course design would be so much better if students cooperated. I would have all my exams be take home, open book (but solo work). But what do I do about the 20% of the students who have zero respect for any sort of honor code and even less fear of any consequences? You don't even have to hunt for the existence of these students. On many cheating related threads on HN, a frighteningly people openly and proudly say that they will and have cheated because its only the degree that matters.

I was also at the teachers side. Unclear rules and situations harm especially students that don't tend to complain and argue.

But back to student time, I had class where we were expected to write super detailed explanation on every homework. If you skipped a tiny step, you got marked down. And then came test. I could answer every single question and did not even stopped to think. When time run out, I was still writing. People complained and the response was that we should not be that detailed on test - but that is exactly what they trained us to do.

The op complained about this: "from classes where I walked into tests expecting that I should be able to answer every question perfectly, to classes where my 70% score on the midterm (of a handful of questions) was considered an excellent result, with no explicit indication beforehand that this would be the case. "

This is sort of guesswork about what you are even supposed to do sux.

> I would have all my exams be take home, open book (but solo work). But what do I do about the 20% of the students who have zero respect for any sort of honor code and even less fear of any consequences?

Don't do home exams, imo. Cheaters always, literally always exists. In some cultures they are effectively normalized and more frequent, in some less. Consequences for cheating in American colleges are quite high actually, they can kick you out. They do not want to be kicked out, but there is no way to catch cheater in their home.

Reply on news.ycombinator.com