What should I do to get into MIT, CalTech, etc?

[twofish-quant]

I absolutely love Georgia Tech. We have a lot of profs who are at the forefront of their respective research areas. One of my absolute favorite profs was Dr. Prasad Tetali who has done a lot of well-known research. On top of that, I have NEVER experience a prof who is just bad at teaching. To me, it really seems that the department does a good job of hiring good teachers.

When I was at MIT in the late-1980's, I had some teachers that were bad and in some cases spectacularly bad. One irony is that because the students are all very good, having an incompetent teacher didn't do that much damage. People grumble and get annoyed, but in the end, people end up learning the material anyway.

One thing that MIT does do which I think is a good thing is that they put the best teachers in lower division core classes. The reason for this is that if you have a bad teacher in an upper level class, the students have already learned the basics so they aren't going to do much damage, and having upper level students go through a bad lecture is outweighed by the benefits of having that student interact personally with the professor in the lab. If you have a bad teacher in a lower division class then it's a disaster because it means that students can't master the basics.

One other thing is that if you have a class of 300 students, you need pretty good administration skills. There are a few professor that I can think of who are decent but not spectacular lecturers, but they are really, really good at making a class of 300-500 students run like clockwork. Also MIT puts a lot of effort and resources into lower division. One thing that MIT does which is one of those simple things that is not so simple is that it hand grades all tests, so that you have a real live human tell you what you did wrong. It's an logistical challenge to hand grade 500 calculus I tests rather than just do multiple choice, but they think its essential.

In a lot of schools it's the reverse. The lower division classes are "weed out" classes and so the school has no particular reason to put good teachers there.

I really like listening to Prof. Mattuck's ODE course; no matter how many times I listen, I think he is really funny and interesting.

Arthur Mattuck is certainly a "personality." One of his good points is that he really, really, really cares about teaching, and when we had a issue in running the course evaluation guide, he was one of the important people we'd go to.

However, while I think MIT is above GaTech

If you've never had a bad lecturer then GaTech is better than MIT in that area.

I think that GaTech and CalTech are on equal footing when it comes to engineering. In fact, GT ranks a bit higher than CalTech in the US News and World Report Engineering rankings.

I don't trust those rankings at all. One problem with rankings is that in order to compare Caltech, MIT, Gatech, you really need to have someone that has taken classes and gone through those schools.

 

[kamikaze1]

Hey, I'm a college freshman who applied to MIT, Caltech, and GaTech last admission cycle, and got in 2 of them (Caltech and GaTech). As an incoming high school student, here are the things to look into.
1) You should get top grades in math and science classes (if not all classes) and SAT without much efforts. Standard hs AP curriculum is terribly easy and provides a really poor foundation, imo, especially if you want to become a mathematician.
2) If you truly love math and science, you need to explore and pursue math and science endeavors outside of classrooms that challenge you and open your horizon of knowledge. Here is some ideas:
- Math and science comps such as AMC/AIME/USAMO, USNCO, USABO, USACO, USAPho etc.
Doing well in these competitions are no jokes, especially for the math series (AMC/AIME/USAMO), you need proper training, background, and talent to do well in math comp., but if ur a math count national qualifier/state champion etc. You are in a good position to do well. If haven't done math count in middle school or haven't heard of math count, then you should try to take the AMC anyway, but don't expect to make USAMO, if AIME, on ur first try. imo, math competitions are the hardest and most stimulating experience, regardless of wat ur interested in science, u should at least take the AMC and do math competitions throughout high school, and win, if not place high on them.
- USABO is the easiest, intellectually, tho u need to spend lots of time learning bio from Campbell bio book. However, making finalist round, and the IBO team is much harder and requires hardcore studying. Many of those have taken advanced college classes. I don't know much about USApho, and others. if ur good at math and have a good physics knowledge, u should do well in it, at least making semifinalist round.
3) Look into Siemens/Intel STS, research based competitions (this should be a good experience too, if u can find a mentor and start ur research project early)
4) Science/math camps
- SSP
- Promys
- MITES/YESS
- RSI (this one is incredibly hard to get in)
- MOSP (top scorers for USAMO) etc
5) Admission matters
- MIT is incredibly unpredictable, since they follow the Harvard model in creating a diverse class. To get in as a science, pure academic standpoint, u need to tank out on the comps i mentioned above and basically become the best in the field. Others got in because of other artsy talents etc.
- Caltech is more predictable: u need to be the most hardcore of the applicant pool in terms of academic standpoint and potential to become a good scientist/mathematicians/engineers. U need basically to be an academic beast, here is another helpful link
http://talk.collegeconfidential.com/california-institute-technology/1182824-suugest-reading-two-threads-before-you-post-chance-me-caltech-type-request.html

Based on what u've posted, u seem to have good credentials and on track to get in those top schools. As u learn more materials and hone ur test taking skills, u should be able to massacre the SAT/ACT reading/writing.

Many got in the two elite schools without any of the things mentioned above (excluding 1). I didn't know about those in my senior year; those are just some idesas, not formulas. If ur in a good and well funded public hs, private schools, u may very well be exposed to the competitions above. I hope this helps. (Edit: i forgot this is a zombie thread, awkward)

 

[twofish-quant]

Have you seen http://prl.aps.org/edannounce/PhysRevLett.100.070001? "If a physicist wants to work in such a collaboration, she or he must be sure to have a tough mind and sharp elbows" Wow! I don't think they run the army like that.

The immense amount of politics in high energy physics is the main reason I stayed out of that field. In part the politics has to do with the nature of the field. HEP experiments are massive engineering projects. One thing that is good about computational fields is that you can spend decades putting together a simulation, but once you have written the code, you can just duplicate it to anyone who wants it, and that makes it possible to have collaborations of three or four people, rather than hundreds.

Watson also has a famous self-indicting book.

One reason that its good to go to MIT so that you see things first hand. It's different to read about something than to have it happen to you personally or people that you know.

Good for Vest! But don't they consider it a form of advertising? Everyone else has their notes on the internet now. Surely MIT has to at least keep up?

Why does MIT need to advertise itself? It's not as if people don't know about it. :-) :-) :-)

Also, decision making at MIT is surprisingly slow and bureaucratic. The reason is that MIT works by consensus, and consensus decision making tends to be slow. Also, MIT doesn't have a very strong tradition of community outreach. Finally, things like distance learning just isn't a high priority for Hockfield. Her vision of MIT is focused on making it a premier research cross-disciplinary institute, and so that gets first priority.

Something else to point out is the fact that MIT is slow helped it. Columbia and NYU were a lot faster to embrace distance learning, but what they came up with (google for fathom) just blew up and they had to shut it down. The thing about OCW is that it still exists.

But one good thing about MIT is that it's part of an general "technology ecosystem". If the next step in distance learning can't happen at MIT, people can just take those ideas and make it work somewhere else.

 

[twofish-quant]

- MIT is incredibly unpredictable, since they follow the Harvard model in creating a diverse class.

It is unpredictable, and diversity is really important to have MIT function. One thing that MIT has found from experience is that if you want to come up with new and original ideas, you just have to have people from different backgrounds. If you have five stereotypical science nerds in a room, they are just going to come up with the standard "science nerd" ideas. So if you want something original, you have to people in the room that are smart but *aren't* stereotypical science nerds.

Also having a diverse student body is important because it gives you more choices. If everyone on campus is a science nerd, they life will suck if you don't want to be a science nerd, but one good thing about MIT is that because there are so many different types of people, you can find a group that you can fit in, or you can decide to do something original and combine the best of all the groups.

The other thing that I liked about MIT is that there is a lot *less* competition than in a lot of other universities. If you have only science nerds in a room, you end up competing over who is the "nerdiest", but what ends up happening at MIT is that because you have lots of people that say "hey, I don't care about being the nerdiest, and I want to be the most artistic."

Something that helped me a lot was that at MIT, I had a whole bunch of teachers and peer pressure to do X, but also a whole bunch of teachers and peer pressure to do not-X, which meant that in the end, I have to figure out for myself what I wanted to do.

The problem with this is that because the freshman class is only 1000 people, you have to reject a lot of people that would just do fine at MIT, and this problem is getting worse over time. Looking at historical admission rates, I'm pretty sure that MIT could double admission rates without reducing student quality, but the problem is that MIT just doesn't scale.

This is why I think MIT faculty and alumni are trying very hard to figure out how to get the MIT experience without getting admitted to MIT, and which is one reason I'm posting as much as I do. If you didn't get into MIT, I want to give you enough information so that you can build MIT or build something better than MIT where ever you do end up.

 

[atyy]

In a lot of schools it's the reverse. The lower division classes are "weed out" classes and so the school has no particular reason to put good teachers there.

Do weed out classes really exist?

A similar rumour is that some universities admit more graduate students than they can take so they have TAs. Then they flunk them out during qualifying exams.

Is there any publicly available evidence for such things? If they exist, aren't they unethical?

 

[twofish-quant]

Do weed out classes really exist?

For undergraduates. They do at University of Texas at Austin. They don't at MIT or William and Mary.

A similar rumour is that some universities admit more graduate students than they can take so they have TAs. Then they flunk them out during qualifying exams.

I can tell you publicly that at the University of Texas at Austin, the *astronomy* department doesn't do this, but I can't vouch for other departments there. If you are interested in rumor and hearsay, e-mail me in private, and I'll tell you what I've seen.

Is there any publicly available evidence for such things? If they exist, aren't they unethical?

This is why it's a good idea to talk to upperclassmen. As far as being unethical, you can argue the point.

One reason that UT Austin has weed out classes and MIT doesn't is that UT Austin has pretty open admissions whereas at MIT people are "weeded out" before they set foot on campus. One absolute firm and total requirement for an admission to MIT is that you have to convince people that you can get through the basic calculus and physics classes.

The other difference is that if you are UT Austin, and you flunk out of physics 101, there are a ton of other majors you can do, whereas at MIT if you can't pass 8.01 you can't get a degree in anything since physics and calculus is a required topic for all students.

 

[Mépris]

One reason that UT Austin has weed out classes and MIT doesn't is that UT Austin has pretty open admissions whereas at MIT people are "weeded out" before they set foot on campus. One absolute firm and total requirement for an admission to MIT is that you have to convince people that you can get through the basic calculus and physics classes.

The other difference is that if you are UT Austin, and you flunk out of physics 101, there are a ton of other majors you can do, whereas at MIT if you can't pass 8.01 you can't get a degree in anything since physics and calculus is a required topic for all students.

While I do think that this makes sense, it isn't explicitly stated anywhere on the website. It can, however, be inferred from that page which says that even if they do offer "humanities" majors, one has to have the pre-req courses to graduate. And the page which says that mathematics up to the level of calculus should've been studied before.

Why is that so? I suspect it's because that might complicate things for them because there'd be more well-prepared people applying. Alternatively it could mean that people who aren't, wouldn't bother applying.

??!?

 

[twofish-quant]

Why is that so? I suspect it's because that might complicate things for them because there'd be more well-prepared people applying. Alternatively it could mean that people who aren't, wouldn't bother applying.

This is some educated guess work, but I much of it is that they don't want to scare off people that have decent math skills, but huge amounts of math anxiety. If you talk about math and physics too much, then you may scare off applicants that would do fine in 18.01 and 8.01, so the philosophy seems to be "give us your application and we'll figure out if you can make it through the general institute requirements."

Something that MIT has obviously tried to do is to broaden the types of people applying, but by having more people applying you run into the problem of too many good applicants, and not enough spaces.

 

[atyy]

This is why it's a good idea to talk to upperclassmen. As far as being unethical, you can argue the point.

One reason that UT Austin has weed out classes and MIT doesn't is that UT Austin has pretty open admissions whereas at MIT people are "weeded out" before they set foot on campus. One absolute firm and total requirement for an admission to MIT is that you have to convince people that you can get through the basic calculus and physics classes.

The other difference is that if you are UT Austin, and you flunk out of physics 101, there are a ton of other majors you can do, whereas at MIT if you can't pass 8.01 you can't get a degree in anything since physics and calculus is a required topic for all students.

What exactly is a weed-out class?

Is it one that is badly taught so that only those who already know the subject can pass?

Or is it one that is well-taught for the stated pre-requisites (like high school algebra - which I believe is the pre-requisite for one version of MIT's freshmen physics), and those who haven't reached the equivalent of 10th grade mathematics flunk out?

It's interesting though, that the most mathematically accomodating version at MIT is given over a longer time, and restricted to 100 students. (I think it's 8.01L of http://student.mit.edu/catalog/m8a.html#8.01l). Perhaps the public schools don't have the resources to have so many different versions of freshman physics?

They also have a version called 8.011 "Designed for students with previous experience in 8.01". Would you know if that's for those who took the course and didn't pass, or is that an advanced version for those who studied the subject in high school but want to do it again to make sure they've got the basics securely?

 

[twofish-quant]

What exactly is a weed-out class?

It's a class in which many/most people who talk the class are expected to fail and drop out. A lot of schools put some tough classes at the beginning to try to reduce the number of students that continue on in that major.

It's interesting though, that the most mathematically accomodating version at MIT is given over a longer time, and restricted to 100 students. (I think it's 8.01L of http://student.mit.edu/catalog/m8a.html#8.01l). Perhaps the public schools don't have the resources to have so many different versions of freshman physics?

What MIT is doing with freshman physics is one of the reasons that I think people should apply.

When I was an undergraduate, there were a few different tracks for 8.01 but they were all "big lecture class, recitation, test, and problem set." What MIT has been doing over the last few years is to totally rethink and redesign how 8.01 is taught, and today most frosh can take several different versions, some of which are taught in very different ways than when I was there.

Now I'm sure that the way that MIT teaches freshman physics will eventually end up to be the standard way at most universities in a few years, but you get it at MIT first.

 

[ladyada]

They also have a version called 8.011 "Designed for students with previous experience in 8.01". Would you know if that's for those who took the course and didn't pass, or is that an advanced version for those who studied the subject in high school but want to do it again to make sure they've got the basics securely?

Heh. 8.011 is a new invention for those who failed 8.01... they used to just teach 8.01 every term, not 8.01 in the fall and 8.011 in the spring. The advanced version of 8.01 is 8.012; it assumes that you have had at least 1D calculus (equivalent of 18.01 or AP Calculus BC) and has a lot more theory and math than 8.01.