How Universities Work, or: What I Wish I’d Known As a Freshman
jseliger.com
jseliger.com
So now I direct people who ask me what this grad student business is to the essay, which has been on my private course sites for a long time but which I've finally decided to make public.
Note: if you have feedback, I'd love to hear it, especially if you're an undergrad (two of my former students read "How Universities Work" and said it made sense; the changes that resulted were relatively minor). If you want to send msgs privately, try seligerj [[at]] gmail --dot-- com.
Thanks for the heads up.
See this comment: http://news.ycombinator.com/item?id=1730253 :
Despite temptation, a beginner's introduction to a topic is not the place to list every permutation of the basic concepts that somebody, somewhere uses, but rather to simply explain the basic concepts. Trying to do both at once results in the student knowing less at the end than when they started; not only do they still not understand, now they're afraid of the topic!
In addition, see what I wrote under the section "Degrees:"
Congratulations: you’re now probably working toward your B.A. (bachelor of arts) or B.S. (bachelor of science), which will probably take four years. If you earn that, you’ll have received your undergraduate degree.
It's easier to write "B.A." in most circumstances than B.A. or B.S. or S.B. or B.F.A. or the various other ways of getting cross the same basic idea.
Your statement that all T.A.s do care a little about teaching is false. I was a T.A. that Did Not Care. There were no consequences to doing poorly, and student grades were controlled across T.A.s so my performance could have no affect on student performance.
Also? Be nice to your T.A. Sometimes I was allowed to wildly inflate grades for students who were enthusiastic/tried hard. Sometimes not. But it can't hurt.
Another random piece of advice I would give is for students to take sections later in the week rather than earlier. My first section was always a disaster, I had no idea what I was doing.
If there is a rubric, follow it to the letter, and don't try to be creative. Rubrics are for larger classes to make the grading even, and they are followed exactly. I knew an A paper from a B paper, but sometimes A papers got Bs and B papers got As because of the stupid rubric (or worse...)
From the essay:
To be sure, there are exceptions: some professors will be hostile or uninterested regardless of how much effort a student shows, and some will be martyrs who try to reach even the most distant, disgruntled student. But most professors are in the middle, looking for students who are engaged and focusing on those students.
Look for:
"assuming you want to graduate on time."
"Go to the professor’s office hours"
"whether the question he has asked is the right one"
I didn't know C very well yet, and trying to learn the algorithms material whilst being distracted by implementation details was simply not worth it. I'd tell my younger self: you're going to screw yourself out of both a grade and understanding of fundamental material.
Computer Science, particularly, has a problem with bright students who managed to blaze through their first-year courses with a 4.0, expending absolutely no effort at all. This is because they already know how to program and are often inculcated with enormously broad knowledge of CS as compared with their peers.
Many of them also know dangerous portions of typical second-year subjects - e.g. just enough to regard their second-year algorithms or operating systems courses as 'oh, this is just like first year, I know this'. Unfortunately, this often refers to the first 20-25% of the course, not all of it, and maybe not even that.
Some of them never quite seem to recover and start regarding the marks they get as arbitrary personal judgements of their character ('you gave me a bad mark because I'm smarter than you and you hate me') or as failings in the system ('you gave me a bad mark because this course is in ML and ML is stupid').
Naturally none of this is helped by the fact that there are copious actual failings in the system. The hours of my software engineering course spent on the waterfall model (back in the early 90s), or mind-numbing lectures on the Seven Deadly Layers of OSI, are hours I'll never get back...
And actually, furthermore deciding what sort of research I should get myself involved in. Does it even matter? -- in line of what jseliger and pg said about choosing majors -- that it doesn't matter, because the more important skills are easily translatable into different fields.
Funny -- I actually wrote another essay called "How to Get Your Professors' Attention" on this subject. Some friends are reading it right now, and I wrote it in response to some of the things I've read on HN.
In any event, if you're trying to get in another lab, I'd try something like this:
1) Find the list of the profs and/or their websites.
2) Look at their research. You might not understand a lot of it, but try to get a sense of what they're doing.
3) Try to read some of their recent papers. You might not understand a lot of them, either, so look at the bibliographies: figure out if there are any seminal papers in the field that are cited that you might have a better shot at understanding. Alternately, try to look for a more accessible source of information on the topic. Profs at your own university might be good for that.
4) Then send them an e-mail saying, "I'm interested in subject X and read Y and Z. I've done A and B to learn more about this subject. Can I drop by your office hours to talk it over with you?"
Almost all profs will have office hours.
5) Part of your goal should be to signal that you're not going to waste the prof's time. One way you can do this is by showing that you've invested some amount of your own time. Profs quickly discover that most people who claim they want to learn don't and that most people who implicitly claim they won't waste their time will.
If you want to see a draft of the essay I mentioned, send me an e-mail.
Does it even matter? -- in line of what jseliger and pg said about choosing majors -- that it doesn't matter, because the more important skills are easily translatable into different fields.
Look for what interests you, and do that. What you major in does matter somewhat -- if you want to be an economist, I wouldn't recommend that you single major in art history, and if you want to be a psychologist, electrical engineering might be less useful than some other fields. But people say things like, "You'll be unemployable if you major in X," or "smart people only major in Y," and in those senses what you major in doesn't always matter.
When in doubt, build your reading/writing and math skills, since those are applicable to nearly everything.
Sure, I'd love to have a look at it. I've just sent you an e-mail (on your gmail address listed on your blog-site).
Thanks!
Oh yeah -- also see pg's undergraduation: http://paulgraham.com/college.html , which is on point here. I think in some essay he points out that if you want to work in a lab, you need to convince the prof that you'll be a net decrease in work. This is harder than it might appear.
Grad school is like entering a medieval guild apprenticeship. The main reason a master would take on apprentices is that they're cheaper than an equivalently motivated and skilled individual. So, your goal is to convince potential supervisors that you're motivated and skilled. The specific research doesn't matter so much as getting a good recommendation.
"But why do I need to learn about matrices? I'm a meteorology major! We don't use matrices at all, it's all just partial derivatives!"
(This is an actual complaint I received when TAing ODE's, which was required by the meteorology major. I didn't even bother explaining how you discretize a linear PDE into a system of ODE's, or why that matters for weather prediction, I just referred her to her department head. )
Universities are supposed to do the research and help students learn things. If the students learn things and scientists do research, everything is all right, I guess. No need to impose a military regime.
However, it might not be true for high schools, for it is more supposed to gather students in one place and keep them from fooling around, than actually teaching them stuff.
This kind of thought process is strange to my way of thinking but plausible; very few teachers I ever had said, "Hey, this is why we're doing what we're doing and why it's important," which is part of the reason I make a point of saying exactly that in the classes I teach. One favorite question: "Why are we reading this, as opposed to something else? What are you supposed to take from it?"
by showing students all they need to know, thus allowing them to skip school and learn the exact same material on their own.
This seems implausible: if students came in already well-prepared or overly advanced for the classes I teach (chiefly Engl composition), I'd be delighted. By now I've taught about ~250 students, and somewhere between 1 and 6 haven't really needed a comp class. Probably only 1. But I can guarantee that more than 6 have thought they didn't need a comp class...
HN has a lot of talk about being self-motivated, and learning on your own, and so forth, which I think is great. But standing in front of a classroom makes me think that the number of people who will really do this is not very great, especially without deadlines and so forth to motivate them (for more on this, see Dan Ariely's book _Predictably Irrational_).
That's not to say the number of very self-motivated people who could acquire the knowledge that school imparts on their own is zero. It isn't. But it's really, really small, which is part of the reason school systems are shaped as they are.
"This transform is actually just a low-pass filter (which you can now derive in two different ways","All that matrix algebra can be used to simplify calculating how two transmission cables interact", "If you add in complex variables, all those calculations you've been doing just got a whole lot easier", etc.; for me, that was always the golden moment when it became clear that someone had put some thought into what I was learning, as opposed to just hitting a course catalog with a shotgun.
Only one professor - one of the signals guys - made this point at the start of the semester. I don't remember the exact problem, but he started the first lecture (after giving out the syllabus) working out on the board a relatively complex problem from another course. Took him about ten minutes and three or four blackboards, and he made a few mistakes along the way. Then he erased all his work, solved the same problem in about four lines using a bunch of symbols/variables/calculations that we hadn't yet been exposed to, and said "in three months, you will understand what these four lines mean, and why they're so much more powerful that what you're used to writing out." Come the end of the semester three months later, he was right - the solution he showed us (he wrote it out again) made perfect sense, and applied in a whole bunch of places.
Having said all that, I'm still not clear why I had to take analytical mechanics/statics in my second semester.
Part of the hazing :) On a serious note, it looks like ABET has relaxed their standards.
> For most of its history, ABET’s accreditation criteria specifically outlined the major elements that accredited engineering programs must have, including the program curricula, the faculty type, and the facilities. However, in the mid-1990s, the engineering community began to question the appropriateness of such rigid accreditation requirements.[8]
http://en.wikipedia.org/wiki/Accreditation_Board_for_Enginee...
New standards (references room-for-interpretation "Engineering Sciences" and not statics, dynamics, nature of materials):
http://www.abet.org/Linked%20Documents-UPDATE/Criteria%20and...
Intro to materials science was especially cool. I knew next to nothing about materials before taking that class, and it opened up an entirely new world to me. My only regret is that the program only allowed me to take one course in materials.
Like having a contest to predict where something rolled down a ramp will end up? School them in the basic equations. Beer, pizza.
Whether or not that should be the case, that seems to be the way that it is.
The Paul Krugman quote is shudder-worthy, but that is a personal bias.
To a certain degree there are computer science programs where the net result is large-scale training in collaborative development. At the university where I worked, there was a graduate CS class where over the course of 10 weeks they rewrote the transportation department's website (enabling real-time tracking of campus busses and wi-fi base stations in all the busses).
However, I was a chemistry/math major as an undergrad, at a primarily research institution. I imagine chemistry departments at primarily undergrad institutions were probably not as involved in the process of "knowledge creation" - that's what REUs were for - and who can say about the humanities and political science?
I don't know if it's always been like this. My undergrad uni seemed to have recently added actual credit for mere attendance. One of the modules reportedly only had a tenth of the people show up (I didn't realise this cause I ditched it after the first two lecture which were exceptionally dull and unenthusiastic). My postgraduate class consisted of a room half full of foreign students who talked all lecture. I don't know if there's cynicism amng academics about their students, or some sort of undying optimism. I suspect they say "unbelievable now lazy they're getting, how lax the standards are, how ignorant..." when we're not there.
And then there's the tutorials: 90% of the time they were a joke. Even the better ones where some active discussion transpired, I wasn't convinced of their actual value beyond entertainment.
I found that this to be vary a lot from department to department. First of all, undergraduate programs across the board are largely about disseminating knowledge, as a foundation to be able to create new knowledge later as a graduate student or researcher. As far as similarity to high school, I found that business school undergraduate classes in particular were very much like high school just on a larger scale: attendance taken, assigned seats, required homework, etc. Classes in the Math or science departments were more likely to leave it all up to the student: grades were based solely on performance on exams. Humanities courses were somewhere in between those extremes.
One thing not mentioned in this piece, as an undergraduate it is much easier to establish a high GPA your first two years than it is to bring up a low GPA as a junior or senior. You don't do this so you can "coast" later, but because your 3rd and 4th year classes in your major are going to be hard, or at least quite time-consuming. If you have a high GPA coming in to your junior year, you don't have to worry so much about the stray required electives you are still finishing off, at the expense of the subjects you are really interested in.
I cannot stand it when a professor says that the best way to get an A is to learn. It shows a fundamental ignorance about their own system of instruction.
If there is anything I've learned about universities, it is that learning is at best orthogonal to getting an A. More often you have to actively sabotage your actual learning in order to get an A.
When you get excited about a subject enough to explore it on your own, you will almost inevitably end up either spending too much time on one part of the curriculum or neglecting the rote memorization and tedious busy work that is necessary to get an A. Classes are just small factories, treating and evaluating every student the same way. It is and ever was a game of points and percentages. If you want "learning" out of this, you need to balance it with playing the game.
Professors have either never known this or have completely forgotten it. They constantly whine about how students seemingly strive to get the least out of their education as possible by taking as many shortcuts as they can. What they fail to realize is that all of their classes are set up as silly little point games, and each student has about 5 or 6 of them. Then they give out so many busy work assignments and so much reading per class that it is impossible to accomplish it all. The students are essentially learning how to prioritize and cut out all of the unnecessary fat in order to maximize their points. They're kept so busy that they don't even know that they're not really maximizing for learning. Learning is not what the professors tell them to do. Learning is not what the professors hold over their heads. They live by points, and die by points. It takes a special student to pull his/her head out of this firehose of tasks and decide to pursue their own education.
So Bob the student comes into class and he hasn't done the reading. Why? Because he has spent 102% of his time on the various homeworks and deliverables for this and other classes -- the stuff that actually has an effect on the bottom line: his grade. He'd sure like to do the reading, but he knows that he can get the assignments done just as well without it, and so it goes to the bottom of the priority list. He will never get around to doing the reading, because the list of grade-impacting busy-work deliverables will never cease.
The professors then complain about the students like Bob, who don't seem to want to actively explore their subject. They complain about the work of their own hands, the inevitable result of their own system. The professor actively molds students to be this way, and then they complain about it.
That is what I wish I'd known as a freshman.
Anyhow, points are a game that even the best students get sucked into playing. We are not passionate about every subject, obviously. So, play the points game for subjects you don't care about, then really try and learn for the subjects you do care about. It will only help your grade, and you might find some friends and supporters along the way too.
For instance: across two projects in two classes, I've clocked over 1000 hours. Both classes had 3 (shorter) projects, and I had three other courses at the same time.
And that wasn't my busiest semester. That one had 21 credits - one of which was an internship - and a part time job. (not recommended under any circumstance, btw)
It took me a semester of physics homework to actually understand vector calculus. I had to write philosophy papers in order to see the beautiful arguments in my head fall apart on the page. I tried to learn a language by immersion alone -- but that left gaps it took hours and hours with a textbook to fix.
Even points aren't a terrible idea: I'm very good at overestimating the depth of my understanding, and I often have to do a certain amount of grinding to uncover the deficiencies in my knowledge. During my undergraduate career, points (when properly applied) were very good at getting me to realize how important some of the pointless-seeming work was.
Good professors will probably give you some "busy work". Whether they're assigning the right work, and providing the right incentives, is another question.
Personally, my favorite setup is what one of my math professors does: recommended homework, which can be submitted to get feedback, and which 1/2 of the test questions are pulled from. If you do it all, it's a lot, but you're not required to do any. Predictably, most people do little, but the people who need it / actually care still get the benefit of homework, and the flexibility of a lack of deadlines.
* Some courses are hard enough, and many undergraduates are undisciplined enough, that a lot of students would end up failing. Of course, when this happens, the professor gets a whole bunch of reviews back telling him that he taught badly.
* Sometimes exams aren't the best setting for evaluation.
2) Yeah, but then you're in an entirely different style of class, and it probably doesn't have a lot of busy-work homework like objectively-testable (if there is such a thing) classes can generate. Though please, point some out to me if you've had any, I'm quite interested in how education is handled :)
It was amazing, as I could devote however much time I needed in order to be able to identify and then efficiently and effectively solve a given problem type.
This was still at the level where it was a lot closer to arithmetic than actual math (even the most complicated problems rested on just one or two methods which we already knew, and all had an objective, known solution), but the flexibility that it offered me was very useful, given the busy-ness of my schedule that year.
The way he had setup the class was that he'd give lecture for three days out of the week, give us two week long assignments, and nothing else.
These limits were extremely reasonable, he went far and beyond in explaining a lot of concepts in easy to understand language, and he was a tough grader. But everyone came away anyway with at least a 3.5 or higher in the class.
Why? Because he was also very very reasonable with the workload. The assignments were anything but busy work - he gave two weeks because he really wanted everyone to absorb the material before tackling the mind crushing assignments.
There was another math class I took in college (that I failed miserably) - it was a senior level analysis class. I think within the first week or two, I was on the list to give a proposed solution to a problem in the problem set. Within two or three seconds of giving a solution, a fellow student shot me down and the professor said in front of everyone that "it was alright, it just meant I didn't understand the material." Never went to class after that.
I'm not saying that makes me amazing or anything, but it's quite possible to fail university out of sheer love of knowledge; in fact far more likely than doing so out of ignorance. Someone who is determined to stay ignorant is going to do very well in university, since their attitude will be realistic; they'll quickly figure out the shortest route to getting a good mark. Actually learning is not the shortest route.
Also, professors have to realize that at least 4 times out of 5 you are just taking the course because of the bizarre credit requirement system, which is in part designed just to funnel students to every bloody department, whether people are interested in it or not.
I never find myself doing this with a CS class. I always show up right before class starts, answer a few questions to show the professor I haven't zoned out in front of my laptop and leave once class ends. Talking to others in the class, they pretty much did the same. What we were covering always seemed pretty set in stone. The complexity of mergesort won't change and its rather easy to see how it differs from quicksort, for a contrived example.
However, when I had International Studies as my second major, I found myself in professors offices a lot more. The politics behind everything at the level we were looking at things on was much easier to find questions to explore than CS was. And even more so with the art classes that I'm in right now. I'll stay after class to work on something, show up before class to talk to others and see how their project is coming along (and theirs is usually in a completely different direction than mine). I'd ask the professors if they knew of any artists who had made similar works that I could look at, etc.
Some do, but my impression is that most don't. I didn't, as I wrote; I knew that most smart people go to college, but I didn't know how college was structured (nor did I know how high school was structured, or what was wrong with the structure of high school; for more on that, see http://jseliger.com/2009/11/12/susan-engel-doesnt-get/ ).
Maybe some students figure out the structure of universities quickly through osmosis, but again, I wasn't one. If this essay hastens the process, then I've done something useful in the world.
So did I. I spent a lot of time during high school looking over options of where to study. When I enrolled, I knew almost as much about the CS department's course offerings and degree programs as the faculty did as well as what courses I would take to fulfill my general education requirements. I was very confused at the large proportion of students I encountered who knew very little about the teachings their departments (and the university) offered. While I now acknowledge that this is the way things are, I still do not understand why. I have not determined whether this correlates with the degree to which students put effort into their own education; this should make for some interesting conversations (I think I'll start with my brother).
I'll posit some reasons:
I didn't know what I wanted to major in or what I wanted to study. This is very common, as is switching majors. It's also hard to evaluate schools because so many factors go into a decision: large or small school, big university or small college, scholarships, tuition costs, geographic proximity, faculty sizes, student culture... the list goes on. And the people choosing are 17- and 18-year-olds, most of whom aren't particularly well formed or knowledgeable about their interests. There's a reason why you're supposed to discover yourself at the university: because most people don't know themselves before they go. There's a reason why the bildungsroman is such a popular genre: most people go through phases when growing up and learning about the world. Maybe you did so earlier than most. I didn't.
I also think that most people are bad at anticipating what their future selves will want. That also applies to me, and my feeling has been reinforced by books like Philip Zimbardo and John Boyd's _The Time Paradox_, in which they talk about time preferences and how those change. One reason I like Paul Graham's "What You'll Wish You'd Known": http://paulgraham.com/hs.html is the hang glider metaphor: you don't know where you'll want land, so stay upwind.
If you _knew_ that you wanted to be a CS major and you stayed in that major throughout your college experience, you're in the minority of students.
I tried and found that there is virtually no negotiating with requirements, even if some are or seem silly. For example, Clark required that students take “science perspective.” In studying my schedule and options, I figured that astronomy was the easiest way out. Considering how useless astronomy looked, I decided to petition the Dean of Students to be excused from it so I could take better classes, arguing that I’d taken real science classes in high school and that I could be more productively engaged elsewhere. The answer came quickly: “no.”
Considering how useless astronomy looked in the context of taking it merely to fulfill a requirement of limited utility, I didn't especially want to take it, but I took it, fulfilled the requirement, and continued.
Astronomy is a lovely subject, like many subjects, but it's not one that I desperately wanted to learn about then, and the class was mostly a waste.
Isn't this true for the entire undergraduate curriculum?
"The biggest difference between a university and a high school is that universities are designed to create new knowledge, while high schools are designed to disseminate existing knowledge"
This is wrong. There are no "new knowledge" in undergraduate level.
I was thinking more about physics. I believe that it takes longer for a student to reach the original research stage in physics.