<p>Statistical significance is definitely too weak a standard in this case. With N approximately 1,000 a 10 point difference might be statistically significant yet I don’t think a 10 or 20 point difference in average SAT scores could reasonably be considered significant. </p>
<p>
</p>
<p>But consideration of rigor can be a factor in the personal preference…that is probably why this thread started and there are over 12 pages of responses. </p>
<p>Guyssss why do we need to have a pissing contest about this.</p>
<p>You can go to any school (any school!) and end up a happy, successful person who makes a big positive impact on the world. You can totally go to UC Berkeley and find yourself kicking ass and taking naps after (and even before!) you graduate, and you can totally do that if you go to MIT, too. You can even do that if you go to a school that is neither MIT nor UC Berkeley. It’s way more important to find a place where you can fit in and be happy and have a support network so that you can take full advantage of the opportunities around you. That’s why top schools care about “match.”</p>
<p>And really, UC Berkeley and MIT are both top schools. Once you get to the top something or other, your future is as golden as can get, especially if you major in a technical field. You’re probably not going to be hungry unless you are hungry on purpose, and you’ll probably get to live in your very own house or apartment and send your offspring to nice schools and have a fulfilling and intellectually stimulating job.</p>
<p>*kicking ass and taking names, not naps</p>
<p>I’m not even going to edit that out.</p>
<p>I agree with most of lidusha’s post – but I don’t think that’s a reason to not discuss rigor. </p>
<p>An opinion from a distance, but based on a reasonably good knowledge of academia:</p>
<p>It really depends on the course sequences that the student selects.</p>
<p>Average rigor of course work at Berkeley < average rigor of course work at MIT (I’m pretty sure about this one.)
Average rigor of course work in engineering at Berkeley (probably) < average rigor of course work at MIT, but by a significantly smaller margin than the average across all majors at UCB
Rigor of many possible course selections at Berkeley > rigor of possible course selections at MIT
Rigor of the most rigorous set of course selections at Berkeley (more or less) = rigor of the most rigorous set of course selections at MIT</p>
<p>The difference in the last category is that a student opting for the most rigorous set of course selections at Berkeley will rarely be in classes with his/her own age cohort–more likely, he/she will be with students 2-4 or more years older.</p>
<p>Then there is the question: What is the point of rigor in course work? If it’s for bragging rights only, there is really no point in it. If it is sought to impress future employers, it makes some sense, but that’s still a relatively weak justification.</p>
<p>I think that the point of rigor is to change how one’s brain works. I believe that this really happens, if a student selects courses that are at the right level of challenge for the student: hard, but not overwhelmingly hard. That could be done either place. It probably takes better advising and more selectivity about the courses at UCB than at MIT, but it’s certainly possible. </p>
<p>It think it’s helpful to look at the program at each school.</p>
<p><a href=“http://student.mit.edu/catalog/m6a.html[/url]”>http://student.mit.edu/catalog/m6a.html</a>
<a href=“http://www.eecs.mit.edu/academics-admissions/undergraduate-programs[/url]”>http://www.eecs.mit.edu/academics-admissions/undergraduate-programs</a></p>
<p><a href=“http://www-inst.eecs.berkeley.edu/classes-eecs.html[/url]”>http://www-inst.eecs.berkeley.edu/classes-eecs.html</a>
<a href=“http://www.eecs.berkeley.edu/Programs/Handbook/section2.shtml[/url]”>http://www.eecs.berkeley.edu/Programs/Handbook/section2.shtml</a></p>
<p>QuantMech, I don’t think the option of taking grad classes solves the rigor difference. As I’m sure you know grad classes aren’t just harder versions of undergraduate classes; they typically have a fundamentally different purpose. Undergrad classes typically provide considerably more motivation while grad classes often focus on being comprehensive and give little motivation. From both my own experience and talking with other people it’s reasonably common for undergrads to take graduate classes and being able to follow all the technical details perfectly well but lose sight of the big picture. This may be better than taking technically very easy undergraduate classes but is certainly less preferable than taking a very technically rigorous undergraduate class on the subject that may be somewhat less comprehensive because it focuses on the motivation. I think you are much more likely to find technically rigorous undergrad classes at a place like MIT where many undergrads are very good and the student:faculty ratio is low enough to accommodate many such specialized classes.</p>
<p>If the above seems vague let me give a concrete example. This spring I took a graduate EECS class in convex optimization. The class required some real analysis and a decent amount of mathematical maturity as many homework problems were proofs. I think this is considerably beyond the technical level of most EECS undergrads at MIT although perhaps not of most math majors. I didn’t know much optimization before but was interested in optimization so I took the class. The other students were graduate students who needed to use convex optimization in their research. As such the class spent a lot of time on very specific convex optimization algorithms and very little time on motivation for the subject. For the graduate students this made sense; they knew why convex optimization was important and cared about learning the best algorithms for whatever class of problem they were working on. For someone like me this is not ideal. I still don’t really know why convex optimization is useful other than l-1 regularization in machine learning and knowledge of 10 convex optimization algorithms instead of like 5 convex optimization algorithms is not terribly important to me (different algorithms perform better or worse on different classes of problems). </p>
<p>
I’m not sure you appreciate how math-heavy the undergraduate computer science program at MIT is…</p>
<p>Sorry, scanned through your past posts and it seems that you’re at MIT right now, at this very moment, as we speak. Are you 6-3? I feel like 6.046 is pretty darn proof heavy. Maybe I’m just bad at proofs?</p>
<p>I am indeed at MIT although I’m course 18. I haven’t taken 6.046 but my impression is that while it requires proofs it does not require the level of mathematical maturity of say 18.100B/C which uses Rudin. Perhaps I’m underestimating course 6 majors but I’m not sure course 6 majors who aren’t doubling in 18 would be able to do well in 18.100B/C which I think is more or less a prerequisite for doing well in 6.253. This may also just be my bias but I think doing proofs with epsilons and deltas require more sophistication than basic discrete math proofs. If you have had different experiences please share though.</p>
<p>Your perspective on graduate courses and their limitations is an interesting one, UMTYMP student.</p>
<p>Some graduate courses are essentially “special topics” courses for grad students who are doing research in a specific area. It looks as though the convex optimization course falls into that category. In many places, there are other types of graduate courses that would provide broader coverage, and some of the perspective that you felt was missing from that course. I don’t know how the grad courses divide up at MIT and UCB, though. It is sometimes hard to tell from the title or brief description.</p>
<p>My understanding is that 18.100 is pretty much the hardest thing ever, yes.</p>
<p>QuantMech, I do agree that some graduate courses provide broader coverage but I’d be surprised if many focused as much on providing intuition as advanced undergrad courses. Maybe the professors I’ve spoken to are anomalies but they often recommended taking undergrad classes even when students had the technical background to do well in grad classes.</p>
<p>Berkeley’s Math H104 uses Rudin also, the same book used by 18.100 B/C. Is seems that Berkeley math students are able, quite like MIT math students, to learn the hardest thing ever. <a href=“Math H104 Home Page”>http://math.berkeley.edu/~ogus/old/Math_H104/</a></p>
<p>My impression is that for UMTYMP, and other students who are highly motivated to find rigor, there is plenty of opportunity at both MIT and Berkeley. For those students who are imagining that it is easy to avoid rigor at Berkeleys COE, and that somehow their C’s or worse at MIT would have been equivalent to B’s and A’s at Berkeley, good luck with that. </p>
<p>Seems a bit strange to compare H104 which is the honors version to 18.100B/C. It seems Berkeley offers 12 sections of 104 every year and one section of H104 while MIT offers two sections of 18.100A, 18.100B, and 18.100C every year. 104 seems to usually use Ross and 18.100A uses Mattuck. It seems believable to me that Berkeley’s honors classes are equivalent in rigor to MIT classes but that hardly proves the schools are equally rigorous. A further point is that 18.100 is far from the hardest thing ever; a large contingent of MIT freshmen skip out of it and took more advanced classes. I don’t think the same is true at Berkeley.</p>
<p>Obviously it would be ridiculous to think C students at MIT are going to become excellent students at Berkeley but I do think for relatively good students at MIT (~20% of the class or so) the academic opportunities at MIT will be considerably better than at Berkeley.</p>
<p>
Really? There are classes here (<em>cough</em> 7.06 <em>cough</em>) that give you Cs for being AVERAGE. At MIT.</p>
<p>There may be rare classes at MIT where the grading is so harsh that a C is average but those are very much the exception. If you define a C student as gpa <3.5/5 then their gpa is either close to a full point below the average gpa or more than a full point depending on the exact gpa. I would be quite surprised if more than 20% of the class had gpas that low. It’s possible these students would be something like average students at Berkeley but I don’t think the difference in abilities is such that someone in the bottom 20% of MIT academically to like the top 20% of Berkeley academically.</p>
<p>Gahhh I wish Mollie had shut down this thread. There is absolutely nothing productive to say here.</p>
<p>MIT ain’t so cool! This other school that isn’t MIT is just as hard as MIT!
Shuttup I sold my soul for 4+ years of being kicked to the ground; MIT is really hard!
Shuttup I sold my soul for 4+ years of being kicked to the ground at another school and it was also really hard!</p>
<p>Also I have not personally taken 18.100 but thanks for throwing shade bro.</p>
<p>There are a lot of things other than “abilities” that affect your ability to perform in classes. I know a few people with GPAs that low. Their abilities are fine. It’s the depression that gets them. I would not be surprised if these kids would have gotten fantastic grades at almost any other school.</p>