Academic Rigor at MIT vs. UC Berkeley

<p>Revealed preference rankings are almost certainly most the most theoretically justified rankings because they are based on actual decisions by people who have large incentives to get the decision right (the current state of data on this leaves something to be desired though). There is a reason why revealed preference rankings are published in top 5 economics journals (The Avery et al paper in QJE) while other rankings are done by for-profit magazines. </p>

<p>

I agree but I think the irrational sense of (non-)self-worth embraced by MIT people which I am thinking about is quite different from the kind you are thinking about. In my experience people at MIT do not tend to think very highly of themselves. Of course there are also healthy, well adjusted people.</p>

<p>

</p>

<p>I’m pretty sure if you have trouble with any one of these you will fail at MIT and fail hard. Every single one of them is essential. I am certain that MIT admissions knows that, especially since some of the admissions officers went to school here.</p>

<p>

</p>

<p>Revealed preferences sometimes say something about the talent of the student body, but not always. Sometimes preferences are driven by a lack of rigor and relative grade inflation among other things. For people choosing between Caltech and Harvard, the grade deflation at Caltech as well as the extreme rigor may dissuade people from going to Caltech.</p>

<p>I’m not from California so I don’t have a good feel for the people in the COE at Berkeley. I am aware that it does have a reputation for academic rigor. And certainly, it is a top destination for grad students or postdocs, on the level of MIT.
In the past, my impression that at a top engineering school like Berkeley, being a top student (top 5-10%) and having excellent scores (700+) would be necessary and sufficient to get in. For MIT, such stats would be necessary but not sufficient. However, in the past decade, Berkeley as well as most other schools have become harder to get into, but again, I would expect admissions results to become more difficult at both schools by an equal degree. </p>

<p>Someone described upthread how their high school had 10 people get into MIT with like 60 get into Berkeley. That was my experience with my own state flagship, which is not quite as renowned as Berkeley but has an extremely strong engineering department.</p>

<p>California does have a lot of talented people, but Berkeley draw from primarily one state with a minority from surrounding states. In comparison, MIT draws from a larger geographic pool. So I would expect the talent pool to be stronger as a result.</p>

<p>One thing I would be interested in would be the number of AIME qualifiers going to Berkeley COE vs. MIT and the number of people who took the AMC exam. Yes, I know not everybody takes it, but the exam has gotten big enough that a lot of the top math/science oriented people do take the exam. However, the brainpower needed to qualify for AIME is much more than is needed to get 800 on the math SAT.</p>

<p>Also, it sounds like @RML is foreign from his description of applying to “American” schools. It’s quite clear that the standards for internationals are much higher than for the domestic pool. So the average international at Berkeley would have to be an extremely strong student. </p>

<p>It is true that revealed preference rankings reflect factors other than rigor. However, they serve as a proxy for the strength of the student body (it is theoretically possible that the stronger cross-admits would prefer college A to college B while among all cross-admits B is preferred to A but there seems to be little evidence for this). Certainly my impression is that as of 2014 most Caltech students are academically strong but didn’t get other offers from top 5 schools as opposed to be really strong students who got offers from top 5 schools and choose Caltech*. Also side point Caltech has experienced rapid grade inflation in the past 5-10 years.</p>

<p>I don’t think data on AIME qualifiers by school exists by a couple of years ago I looked at where class of 2012 high school students who got 7+ on the 2012 ARML (n=35) went to college and found about half go to MIT. From the MIT class of 2016 facebook group I estimated MIT got about ~25% of class of 2012 seniors who qualified for USAMO in their senior year (n=102). Putnam scores are also in line with these results. Obviously these thresholds are much higher than AIME qualification but they show that at the highest levels of mathematical ability there are many more such students at MIT than Berkeley. </p>

<p>*Caltech is a confusing case for revealed preferences. The Avery et al study conducted 10+ years put Caltech second while more recent data shows it losing cross-admits badly to MIT, not taking many students from Stanford, and ranked outside the top 5 in Parchment’s rankings. I think the most likely explanation for this is (i) noise in the Avery study (n~3,000) and (ii) Caltech decreasing in desirability in the past ten years (also seen in admissions rate which used to be lower than MIT’s and is now somewhat higher). I admit this is still a puzzling case as the rest of the data points points to a clear and consistent top 5 (HYPSM) in terms of revealed preferences. </p>

<p>@"UMTYMP student"‌ I don’t think you’re understanding why it is not a proper measuring stick for academic rigor. In your theoretical example, you state that “all” of the stronger candidates would choose the stronger school and not the weaker school. That does not reflect reality at all, the admitted applicant pool would never act in unison like that. But let’s be frank, there are numerous reasons why a student would choose school A over school B, and vice versa. It is not uncommon (I can point to numerous instances on this board) where an individual would be deciding between a stronger and weaker school due to factors such as location, culture, etc and not automatically choose the weaker school. </p>

<p>Preferences are important for economic publishings because consumer preferences are VERY important to economics. However, it is well known and accepted that preference is not always indicative of superior quality. By saying that they have a large incentive to get the decision “right”, what you are saying is that the right decision for them is based off of academic rigor, and little else. That does not reflect reality at all. If you truly think that cross-admit data is a indicator of academic rigor, I would suggest that you read up on how Princeton is reevaluating it’s grade deflation policy (which produces a highly rigorous environment) due to decreasing yield rates, most notably to Yale.</p>

<p>Let’s get to the point, can you provide clear evidence that shows that MIT students are substantially (much) stronger than Berkeley students? We’ve already agreed that they are on average stronger than their west coast peers. I think you keep on trying to prove what numerous people in this thread have already agreed to while spewing data that is irrelevant that you truly don’t understand.</p>

<p>@collegealum314‌ Agreed. But I would like to note that state schools being what they are, tend to admit an overwhelming amount of students from their state. I think that this can explain why some schools have a large amount of students that admit Berkeley students compared to MIT. These schools are most likely CA schools, while schools outside CA probably have a very small number (if any) amount of admitted students at their individual schools. The undergraduate student body at Berkeley is also much large, more than 5 times the size.</p>

<p>I said nothing about all students acting in unison. If a group prefers college A to college B then most but not all students from the group will choose A over B. I agree there is substantial heterogeneity in individual preferences. That is why the dominance of MIT over Berkeley in getting cross-admits is so striking. The vast majority of students whose idiosyncratic preferences would favor Berkeley are choosing MIT. You also neglect the mechanism that more desirable to students leads to a stronger student body which leads to increased rigor. I don’t think cross-admit data is a perfect indicator of academic rigor but I think it’s (a) a decent proxy and (b) shows at the undergrad level MIT and Berkeley are not peer schools. </p>

<p>I think (i) the difference in SAT scores and (ii) that almost all students who get into a top 5 school don’t choose Berkeley is pretty clear evidence that Berkeley students are substantially weaker. The math competition data I posted earlier also shows that in terms of math competition ability MIT has far more top students than Berkeley. Math competitions obviously aren’t everything but they are a decent proxy for math ability.</p>

<p>You did. When you referred to “strong cross-admits” you did not say “some” which means that you referred to them as a group. Now you’re just saying non-sense for the sake of doing so. Please tell me how you took into account that Berkeley is a state school and how it heavily skews yield. The “analysis” (if you can even call it that) you presented from a one-sided point of view (does not taken into account Berkeley) is laughable to anyone who has taken a basic statistics course. Please present your analysis and demonstrate that a Facebook group meets a criteria for a statistically viable sample. If you truly think that that mechanism is so strong, please look at the grade deflation policy for Princeton that I have already mentioned and the controversial grade inflation at Harvard which even their own faculty members says is an indicator of insufficient academic rigor. I would think that Harvard has a highly desirable applicant pool and student body. Have you looked up any rankings by subjects for both schools? Or have you ignored them?</p>

<p>You think wrong. There is a difference in SAT scores but is it substantial? Please quantify what you believe is substantial. Even the Collegeboard agrees that SAT scores are an insufficient indicator of college readiness, yet you choose to ignore this repeatedly. No. And again, cross-admit data is a poor indicator of academic rigor for reasons that numerous people in this thread have stated time and time again. If you want to look at awards, you’ll see that Berkeley alumni (alumni being degree holding) have won 30 Nobel prizes while MIT has won 30. I think that my data above is not only more accurate than data from a FB group, it is much more telling of capability than a math competition. </p>

<p>Using your logic, that cross-admit data is an indicator or sorts of academic rigor, according to Parchment, Pomona College and Harvey Mudd are more academic rigorous than U Chicago and Columbia.</p>

<p>

</p>

<p>I think a reasonable human being can parse this as a statement about a group’s trend rather than every single individual in a group.</p>

<p>Why don’t you present evidence for your side?</p>

<p>Why does Berkeley being a state school heavily influence yield? I wasn’t taking a sample from the Facebook group; I counted the number of such people in the Facebook group. This is an estimate because not all MIT students would be in the Facebook group and some admits who did not go to MIT may still have been in the group (I did this well after May 1st). I don’t think rigor is just defined in difficult of earning grades. Harvard is not particularly grade inflated compared to peer schools despite popular conceptions. </p>

<p>There is a difference between comparing SAT scores of individuals and SAT scores of large groups. I think 100+ points is certainly a substantial difference when looking at groups and a difference of like 50-100 is borderline substantial when looking at groups. There is obviously much more noise when looking at individuals. </p>

<p>The Nobel data includes holders of graduate degrees. It is also based on people who attended those schools 30+ years ago. There is no reason to expect rankings to remain static over such long periods. </p>

<p>Regarding the Parchment data, I don’t think the Parchment college rankings are executed particularly well even if they have a sound theoretical basis. The rankings are incredibly noisy year-to-year outside the very top (HYPSM have always been in the top 6 and composed the top 4) so it seems best to take a multi-year sample and to give little weight to small differences. Lastly, indicator does not mean draw large conclusions from small differences. Berkeley gets less than 5% of cross-admits from MIT while these schools seem roughly similar in terms of cross-admits. Also Harvery Mudd being more rigorous than Columbia or Chicago isn’t that unreasonable. </p>

<p>@PiperXP I think that it is reasonable to assume that when some says “that the stronger cross-admits would” that they are stating that if you are a stronger cross admit you will do X.</p>

<p>I have presented evidence and refuted your side. I think it is reasonable to assume that someone read someone else’s posts in it’s entirety before responding. No only have I shown evidence that shows why cross-admit data is not a good indicator (see my reference to Parchment), why SAT scores are not a good indicator of college readiness (as the College board themselves has said), evidence of preference and academic rigor with respect to the applicant pool by citing Harvard, Princeton, and Yale. You have failed to show that the difference is substantial or even to answer the simple question of what you believe is substantial. </p>

<p>The evidence that your side has shown is poor. If you are going to use cross-admit data to justify that an institution is substantially stronger, then you are saying that Harvey Mudd and Pomona College are much stronger than Columbia and U Chicago. Laughable. If you are going to use SAT data, you are going to have to go against the policy revision of the Collegeboard and how it is not a good indication of college readiness or future college performance. Additionally, you failed to quantify or show that the difference between the two with respect to SAT scores is substantial. Furthermore, the data that the poster I was responding to presented is laughable and statistically questionable. It is a “factual statement” as you say, that Noble prizes are are a substantially more rigorous test of the student body than a math competition.</p>

<p>This is all a repeat of what I’ve already posted. Try reading it first before responding. I am asking for evidence from your side because what you have presented so far is weak and questionable. So let me ask this of you, why don’t you present evidence for your side that holds water?</p>

<p>I think that just looking from your post history you have a dislike of the UCs. However, regardless of your feelings I think you should look at the facts. This is not an attack on your character, but a legitimate observation and concern that has been brought up by other posters.</p>

<p>

</p>

<p>It’s not laughable. U. of Chicago is one of the schools I’d put up there as being in the top 5 in terms of rigor, but in my opinion Harvey Mudd is on par. My guess is that Harvey Mudd is more rigorous than Columbia. If I was to guess which two liberal arts colleges had the most difficult curriculum, it would be Swarthmore and Harvey Mudd. </p>

<p>Pomona is a pretty elite liberal arts college, and I know they get some pretty smart people, well beyond what is typical for Claremont Colleges, but I wouldn’t say it is more rigorous than U. of Chicago. </p>

<p>I do think the liberal arts college element may throw off the interpretation of cross-admit preference.</p>

<p>If you look at the rankings and exclude LACs, you’ll still see questionable findings if you use it as a measure of academic rigor. BYU would be more academically rigorous than Cornell or UCLA using that logic. Additionally, CMU is ranked below many other schools at #115, even a community college at #95. I don’t think that the logic holds. </p>

<p>@mike011x - Claiming to have refuted something does not mean you successfully have done so, and you have not presented any positive evidence for your side. Unfortunately, we cannot get every application that’s crossed MIT and UCB for the past few decades, so we make approximations based on available evidence. Show me something concrete that you believe is better than what we’ve offered.</p>

<p>

</p>

<p>Theories often have holes in them–in particular, there may be specific cases for which they aren’t accurate because confounding factors may dominate. One of your examples is BYU, which someone doesn’t apply to unless they are Mormon. Clearly, this may make the school preferred over some ivy leagues to those who apply. </p>

<p>You can’t figure out what is going on simply through number crunching.</p>

<p>So let’s be perfectly clear just because something (revealed preferences rankings) are an indicator of something else (academic rigor) does not mean that in all cases if college A is ranked higher than college B according it to the indicator it is more academically rigorous. Given that there are many indicators one would think this hardly needs explaining but clearly it does. This is particularly the case when the indicator is noisy like the Parchment rankings. For whatever reason the Parchment rankings are not particularly stable year-to-year and obviously we shouldn’t draw too many conclusions from such noisy data. However, what is perfectly clear in the Parchment rankings, the Avery rankings, and looking at MIT and Stanford’s internal data is that Berkeley gets almost no cross-admits against them. Because Berkeley loses so badly by this metric even if it is a decidedly imperfect one it is clear evidence on the strength of the student bodies. This is not absolutely conclusive evidence but gives it fits with other evidence we can be reasonably confident in our conclusions.</p>

<p>@PiperXP‌ I commend your ability to be so evasive. If you read my previous post, just a little bit you’ll see what evidence I have presented and where I have shown evidence that supports my side and shows flaws in it. I will include it verbatim at the bottom of this post. Instead of avoiding my posts and constantly asking me to present evidence (which I have) which you ignore, why don’t you try and defend yourself? Additionally, it would be useful if you could actually respond to the criticisms that I have. </p>

<p>@‌umtymp I’ll let you ponder this, how would a cheaper large public university with policies that favor students from that state effect yield? You’re using them as a sample, as a smaller group to reflect the larger student body of MIT. I think you need to review your statistics. Your data is weak and using it to make an estimate for such a certain claim makes me question your credibility. Furthermore, you say that MIT students have X amount of Y, while you do not specify Berkeley’s. How can you measure something against something else without data for the other side? Again, refer to the reevaluation of the deflation polices of Princeton and how they have been lossing cross-admits to Yale in a substantial way. Is Princeton not a peer of Harvard?</p>

<p>If you look at percentile rankings for SAT scores you’ll see that the difference is not significant, especially when you go higher. Even the difference between a score of 800 and 720 is a percentile difference of 4 points. I think that the average person would agree that the person in the 99th and 95th percentile are not significantly different in capability. Keep in mind that I’m using these numbers as an example, I am not saying that these percentile represent each school. But the logic is the same. </p>

<p>The scores and standards of a Berkeley student (like colleges overall) have increased in the past 30 years. If anything, this example would be even stronger due to the higher caliber of students today. But the point is, that we can look at many prestigious awards and see the same thing.</p>

<p>Now you’re just refuting your own evidence. The cross-admit data is calculated to be statistically significant to 95%, you can refer to their website to verify. Their data is sound and I don’t think that you can now turn around and say that cross-admit data is now flawed. I don’t think you understand that how a debate works. The data is not only valid if it benefits you, especially when it is mathematically sound. They did not use theoretical data, they simply used cross-admit data. They are not saying that the have a theory that it is indicative of academic rigor like you are. And if you want to talk about “noise in the data” and how it changes from year to year, you should consider that you were just against any measurements that takes into account extended periods of time, such as Nobel prizes. It’s as if you’re arguing against yourself. The parchment data uses the most recent data. And the key word is substantial or “much”. </p>

<p>"@PiperXP I think that it is reasonable to assume that when some says “that the stronger cross-admits would” that they are stating that if you are a stronger cross admit you will do X.</p>

<p>I have presented evidence and refuted your side. I think it is reasonable to assume that someone read someone else’s posts in it’s entirety before responding. No only have I shown evidence that shows why cross-admit data is not a good indicator (see my reference to Parchment), why SAT scores are not a good indicator of college readiness (as the College board themselves has said), evidence of preference and academic rigor with respect to the applicant pool by citing Harvard, Princeton, and Yale. You have failed to show that the difference is substantial or even to answer the simple question of what you believe is substantial.</p>

<p>The evidence that your side has shown is poor. If you are going to use cross-admit data to justify that an institution is substantially stronger, then you are saying that Harvey Mudd and Pomona College are much stronger than Columbia and U Chicago. Laughable. If you are going to use SAT data, you are going to have to go against the policy revision of the Collegeboard and how it is not a good indication of college readiness or future college performance. Additionally, you failed to quantify or show that the difference between the two with respect to SAT scores is substantial. Furthermore, the data that the poster I was responding to presented is laughable and statistically questionable. It is a “factual statement” as you say, that Noble prizes are are a substantially more rigorous test of the student body than a math competition.</p>

<p>This is all a repeat of what I’ve already posted. Try reading it first before responding. I am asking for evidence from your side because what you have presented so far is weak and questionable. So let me ask this of you, why don’t you present evidence for your side that holds water?</p>

<p>I think that just looking from your post history you have a dislike of the UCs. However, regardless of your feelings I think you should look at the facts. This is not an attack on your character, but a legitimate observation and concern that has been brought up by other posters."</p>

<p>@"UMTYMP student"‌ That is obvious, and I think that you should read the portion of your post that I quoted where you clearly state so. If you didn’t mean to say it, then don’t. According to you, using any measurement that involves extended periods of time, such as Nobel prizes, are not valid and thus measurements with the most recent data should be used instead. Then you say that parchments data is not valid because it changes over the years due to it being updated every year with the most recent data. You are contradicting yourself and doing do only when it is in your favor. How is this reasonable? Then you go on to use parchments data after criticizing it. Wow. </p>

<p>However if you are going to say that the logic of using it holds because of the comparison of MIT to Berkeley in the ranking due to one being lower and one being higher, you would also be arguing that that assumption would hold for other examples. Otherwise your logic would not be valid. Using your logic, explain to me how Villanova U at #90 is more rigorous academically than CMU which is ranked at #115. Keep in mind that your argument that cross-admit data is that if school X has a higher cross-admit ranking, it is more academically rigorous. </p>

<p>@mike011x - You still don’t have positive evidence for your side. This is a very simple request, but you are the one being evasive about it :)</p>

<p>I think PiperXP and mike011x should go to binding arbitration.</p>