Hacker News Reader

Let me gently disagree with you. First, much of what she comments on (TAs, etc.) happens in recitations or office hours, which are the time to go over things until they're clear. (She mentions this: "Being hopelessly lost in lab"). Labs are typically smaller and are often taught by less-experienced people - grad students or even senior undergrads.

Second, while it's possible to take it too far, I love it when students are forthright about not getting it, even in the primary lecture session. First, because for every student who asks the question, there are a dozen sitting there being quietly confused -- and often feeling like crap about it (c.f., impostor syndrome). Second, because it gives me an excuse to explain the same concept in a different way -- which may reach some of the other dozen students who incorrectly think they get it, and won't figure out they've missed something until they sit down to do the homework. And third, because that's what learning is about - learning the things you don't know. Any experienced teacher should be able to defer when the questions are getting in the way of the material, and _holy cow_ is it better to have a classroom of students asking you to explain something than to have a bunch of drooling post-lunch zombies.

Your point about when to study & learn things depends a lot on the class. In my intro class, I don't expect the students to have done the reading beforehand, because we don't have any reading. I expect them to have worked through all of the stuff resulting from the previous lecture, however. Other classes operate in the way you mention. But it varies highly from class to class.

The author was taking personal responsibility by being forthright about what she was having a hard time understanding. It's harder to do than it seems.

I think we agree but differ in how we perceive the author's level of understanding.

I didn't mean to imply that there should be no questions in lecture or that students should understand everything by lecture. I also should have commended the author for getting TA help, though I've noticed that students do this as a replacement for studying rather than a supplement--which can fuel learned hopelessness (i.e. "I'm trying so hard and just can't get it").

But I still learn towards bad study habits here. The author felt so helpless that they screamed "I DON'T UNDERSTAND" in the middle of lab. I think that if you're at that point then there's something deeper going on there then just needing more help/clarification and in my experience that has to do with poor study habits.

It's interesting that you don't assign reading before hand in your intro class. If I were taking that class, I would read ahead any way. What advantage do you think your students get by not having an opportunity to look over terms and concepts before you discuss them in lecture?

It's also possible that it was the scenario others have mentioned -- there's a very wide range of incoming abilities in intro cs classes, from complete newbies to high school students who could out-program half of the HN crew. (Seriously - I had one working for me for the summer before he started as a freshman at Berkeley who was the equal of many Google engineers when it comes to C++. Yes, quite unusual, but there's at least one such person.) As a consequence, it can be easy to lose 20+% of the class, and when they get lost early in intro classes, they stay lost, because the concepts build upon each other so rapidly.

Oftentimes, the students don't even realize they're lost until a few weeks later, and by then, it's a big problem. I've seen this a lot - a large amount of our TA effort is dedicated to helping bring those students forward once the "oh shit" moment happens.

Why no reading:

(a) Haven't found a book worth its salt for the class. We have lots of optional texts that they can use, and we encourage students who want to read ahead to peruse How to Think Like a Computer Scientist (python).

(b) It's mostly a structure thing. 15-112 (www.cs.cmu.edu/~112) is at present structured as "go to class, get the big picture view and some details, and then go practice the living daylights out of it afterwords". It's very effective for students new to CS. I can't claim credit for it - my many-awards-winning-astoundingly-beloved colleagues Mark Stehlik and David Kosbie created the beast. The core of it is that students just learning CS need to get the "muscle memory" down so that when you tell them to implement merge sort, they don't spend time puzzling out how to append an item to a list. So the focus of the class is a fairly daunting amount of practice, in the form of exercises, homeworks, and quizzes.

It's a truly brutal class, FWIW, but the students come out of it having accomplished something really amazing. Here's a vide of some of the project highlights from last semester - https://www.youtube.com/watch?v=fkiK_cqSOwI . Keep in mind that these are projects by people who, in many cases, have never programmed before the start of the semester. Full project gallery is http://www.cs.cmu.edu/~112/gallery.html

With such a wild variance of incoming skill levels, what is the way to make lectures efficient? I know this isn't a new problem but I see it frequently in training / classes in the industry where things are 10-100x too slow for some people, while others are lost. There is clearly a failure in assigning people to the correct course level (or offering courses at appropriate granularity), but the more scalable thing I see is offline learning where you first check out slides, then check out accompanying documentation / videos, then check google or IRC, and as a last resort you launch the nuclear option of stopping your life for 2-16 hours to sit in a room and learn synchronously. Is anywhere innovating like this?

I seriously have no idea what her point is. The exact same thing (people having no idea what is going on, lots of them too afraid to ask) happened in maths and biology class. That is what tutorials were for.

>If you don't do the reading and study the material you need to know before the class it's covered in, then of course you won't understand what the professor is on about and of course you'll fall behind.

I don't think I've ever done this aside from explicitly assigned readings for something like an English class. I've always either found the explanations given in-lecture on a topic the first time to be fully sufficient, or had to fill in a few things by reading the book after class.

Your point about responsibility of the student is fair.

On the other hand: https://www.google.com/webhp#q=define%20pedagogy

The teachers have a responsibility to teach, in the face of varying readiness for their teachings in their students. Any asshat can teach people who already know the material, but good teachers teach those who don't.

The responsibility goes both ways. If I'm teaching someone and he/she fails to show at least the effort to look up the basics and achieve them, I'd find it very condescending.

Sometimes like this the person just have to be honest with themselves if this path is really what they want and what are they going to sacrifice to achieve it.

While in computer science I don't have an issue to learn stuff by myself. In math and statistics, I need the context first before I can grasp what a written text about math and statistics really means.

A teacher can fill this gap. Yes as a student, you should do your best, unless classes are splitted by level you will always get this tension.