Luck only gets you into MIT after you pass the cut of it deciding you can do the work, most especially a term of the calculus past the AP BC sequence, and calculus based mechanics and E&M. When I last had figures (since then I gather applications have increased), about 13,000 applied, and about 3,000 were judged to be able to do the work, from which MIT had to extend enough offers to get a class of ~1100.
MIT does a good job of this, as a rule students who choose an appropriate major make As and Bs, unless they're having personal problems.
However these sorts of smarts have no bearing on the failures experienced by this guy; heck, he would have been better off staying longer and rubbing shoulders with people who've had success in business. There's plenty of opportunity for that, from the professors who are given one day a week for their own business pursuits to all sorts of formal entrepreneurial stuff.
But it's not something MIT cares about in admissions, besides looking for evidence you can do projects, which has no bearing on projects producing stuff people will buy; note this guy did not fail from an inability to make something work per se.
Nope, I've carefully looked at MIT's history of grade inflation (the info is online), and there's been none except for a "'60s" period that coincides with the draft, a widespread phenomena of the time. Heck, since I attended in the '80s, they've reduced the pass-fail period from your entire freshman year to just your first term.
What's happening is that MIT very carefully matches its student body to the demands it places on them, something that's worked out well since the post-Sputnik reforms in K-12 science and math education (for more historical perspective, when it started, post-Civil War, a large fraction of students would only be mastering the calculus at the end). There's many levels of oversight including the "outsiders" in Visiting Committees, feedback in the system, etc.
And MIT really cares about undergraduate education (which I gather is all too rare except at "low end" higher education where the school has no pretenses about doing a lot of serious research; one of my chemistry TA came from such a school). All classes are taught by tenured or tenure track professors, with exceptions that prove the rule such as SF author Joe Haldeman, and professors don't get tenure unless they're adequate at teaching.
I once witnessed the EECS department head make a professor read every student evaluation (i.e. students providing feedback) for a course he'd just taught, all but one of which were negative (and the positive one was a special case), and at the end told him he'd never again be allowed to teach that particular course. Etc. etc.
ADDED: MIT tries very hard to be a meritocracy, and since it's focus is on "you can't fool Mother Nature" engineering and science, that's practical.
Semi-disclaimer: I have a lot of ties and history with the EECS department, but being an EECS major is not one of them, I was science track.
> Nope, I've carefully looked at MIT's history of grade inflation (the info is online)
Citation? Every source I can find shows that a) they haven't released average GPAs since 2000 and b) from 1950-2000 they inflated grades at or above the average national rate.
Source was as I remember The Tech, MIT's student newspaper, but I can't find it right now, it was a raw data series. This '75 article starting on page 5 http://tech.mit.edu/V95/PDF/V95-N11.pdf covers '62-'72, but the data I looked at went quite a bit further. If you're really interested I can try harder.
Note that you should throw out results from the '50s and a bit further, when science and especially math education was poorer and MIT had to accept and weed out students. If one was really serious about this, you'd have to find the transition to when they could send acceptances to students they were quite sure could do the work, which had long been the case when I was got in in '79.
Can't say anything based on data about this century; do wonder why they stopped releasing them, but there are reasons that wouldn't touch on grade inflation per se, e.g. the dot.com crash resulted in a great discontinuity for the EECS, which lost more than half its students, after having 40% of them for decades. That resulted in the panic which replaced SICP/6.001 with harder introductory courses (it's even less a place where a pure CS type can slide by on two pure EE courses after the first required course), and it was some time before EECS enrollments returned (and not just at MIT, e.g. Stanford experienced this).
Hmmm, I'm not at all sure what that period might have done to the outside of EECS student body, but I could see many mathematically strong students simply not going to MIT, and my source on what's going on, who would definitely know in at least the School of Engineering, hasn't mentioned any grade inflation problems, I'll ask if that's changed.
There's also one discontinuity before this century, when a professional admissions director actually followed the faculty's official guidelines/rules/whatever on who to admit. There are two classes that were much less strong, this was fixed when e.g. physics faculty members made a big fuss and admissions got realigned with reality.
As for "every source", there's plenty that talk about MIT "suffering" from grade deflation. Actually a problem with others interpreting MIT GPAs vs. schools with notorious grade inflation, pre-meds are understandably very concerned about this; search on mit grade deflation.
Yes, there are plenty of articles discussing MIT and grade deflation. Unfortunately, none of the ones I could find have any data - its all anecdotal from students. The hard data I did actually find (http://www.gradeinflation.com/) shows grade inflation at MIT matching that at Harvard (the media's favorite grade inflation target) almost perfectly from 1950-2000.
I never thought this before, but now that I've actually looked at what little data is available, it looks an awful lot like the MIT grade deflation thing is a myth.
Ah ha, this site is where the data series is, specifically: http://www.gradeinflation.com/MIT.html and my memory about The Tech is from where they got a lot of their data.
So, you're telling me, the GPA going down 1/100 of a point in the quarter century of 1976-2000, is evidence of grade inflation???
If you can seriously look at this time series, accounting for the two factors of better post-Sputnik math and science education and Vietnam War avoid the draft, or simply limit yourself to the period after both of those ('73-75 for the draft, by which time the upper level post-Sputnik reforms would have hit the applicants), then factor in that freshmen went from pass/fail to grades in their second term during this period, and say MIT has statistically significant grade inflation, we have no basis for a discussion.
If we're allowed to pick and choose our time series, the quarter century before that MIT saw inflation at almost double that of the grade inflation poster child, Harvard.
Yeah, you can rationalize based on outside events, but without real data on how much of the change is actually attributable to those things, you're just fitting facts into your worldview. How do we know there aren't similar events in the quarter century you've chosen that are forcing deflation?
Well, some high ranking schools are said to admit a subset of minorities who can't do the work, which has a ripple effect on down. E.g. look at the "minorities as mascots" thesis, and I think Thomas Sowell has written on this.
MIT, for a bunch of reasons I've mentioned, plus intense self-selection in just deciding to apply in the first place, doesn't have this problem. Certainly didn't in the dozen years I was a member of the community, '79-'91, I never met a student who couldn't do the work.
Now, there are students who don't succeed because of financial issues (my problem, and one of at least one close friend), and personal issues (more than a few), but these are not things the admissions office can do as good a job in screening for.
Meta to the article itself but this SumoMe stuff is incredibly annoying. It breaks the flow of reading. There are highlights on here which end in the middle of a word [http://i.imgur.com/6zQjwvW.png]. And now I just had an email-newsletter popup occur halfway through reading the article.
I'm one of those guys who randomly clicks or highlights text while reading, and it's dead annoying if doing clicks on a text with the normal text mouse cursor causes an action!
Please, website owners: provide visual cues if the behavior of your website differs from normal websites.
I'm curious to hear more about the 'reality distortion field'. You say 60-80% of people failed to use the tool - what was the basis for continuing on? Did you have a plan to mitigate the visits to other sites? When was the next time you tested and what were the results?
22 comments
[ 2.9 ms ] story [ 45.9 ms ] threadMIT does a good job of this, as a rule students who choose an appropriate major make As and Bs, unless they're having personal problems.
However these sorts of smarts have no bearing on the failures experienced by this guy; heck, he would have been better off staying longer and rubbing shoulders with people who've had success in business. There's plenty of opportunity for that, from the professors who are given one day a week for their own business pursuits to all sorts of formal entrepreneurial stuff.
But it's not something MIT cares about in admissions, besides looking for evidence you can do projects, which has no bearing on projects producing stuff people will buy; note this guy did not fail from an inability to make something work per se.
This is grade inflation, not a badge of pride to MIT.
What's happening is that MIT very carefully matches its student body to the demands it places on them, something that's worked out well since the post-Sputnik reforms in K-12 science and math education (for more historical perspective, when it started, post-Civil War, a large fraction of students would only be mastering the calculus at the end). There's many levels of oversight including the "outsiders" in Visiting Committees, feedback in the system, etc.
And MIT really cares about undergraduate education (which I gather is all too rare except at "low end" higher education where the school has no pretenses about doing a lot of serious research; one of my chemistry TA came from such a school). All classes are taught by tenured or tenure track professors, with exceptions that prove the rule such as SF author Joe Haldeman, and professors don't get tenure unless they're adequate at teaching.
I once witnessed the EECS department head make a professor read every student evaluation (i.e. students providing feedback) for a course he'd just taught, all but one of which were negative (and the positive one was a special case), and at the end told him he'd never again be allowed to teach that particular course. Etc. etc.
ADDED: MIT tries very hard to be a meritocracy, and since it's focus is on "you can't fool Mother Nature" engineering and science, that's practical.
Semi-disclaimer: I have a lot of ties and history with the EECS department, but being an EECS major is not one of them, I was science track.
Citation? Every source I can find shows that a) they haven't released average GPAs since 2000 and b) from 1950-2000 they inflated grades at or above the average national rate.
Note that you should throw out results from the '50s and a bit further, when science and especially math education was poorer and MIT had to accept and weed out students. If one was really serious about this, you'd have to find the transition to when they could send acceptances to students they were quite sure could do the work, which had long been the case when I was got in in '79.
Can't say anything based on data about this century; do wonder why they stopped releasing them, but there are reasons that wouldn't touch on grade inflation per se, e.g. the dot.com crash resulted in a great discontinuity for the EECS, which lost more than half its students, after having 40% of them for decades. That resulted in the panic which replaced SICP/6.001 with harder introductory courses (it's even less a place where a pure CS type can slide by on two pure EE courses after the first required course), and it was some time before EECS enrollments returned (and not just at MIT, e.g. Stanford experienced this).
Hmmm, I'm not at all sure what that period might have done to the outside of EECS student body, but I could see many mathematically strong students simply not going to MIT, and my source on what's going on, who would definitely know in at least the School of Engineering, hasn't mentioned any grade inflation problems, I'll ask if that's changed.
There's also one discontinuity before this century, when a professional admissions director actually followed the faculty's official guidelines/rules/whatever on who to admit. There are two classes that were much less strong, this was fixed when e.g. physics faculty members made a big fuss and admissions got realigned with reality.
As for "every source", there's plenty that talk about MIT "suffering" from grade deflation. Actually a problem with others interpreting MIT GPAs vs. schools with notorious grade inflation, pre-meds are understandably very concerned about this; search on mit grade deflation.
I never thought this before, but now that I've actually looked at what little data is available, it looks an awful lot like the MIT grade deflation thing is a myth.
So, you're telling me, the GPA going down 1/100 of a point in the quarter century of 1976-2000, is evidence of grade inflation???
If you can seriously look at this time series, accounting for the two factors of better post-Sputnik math and science education and Vietnam War avoid the draft, or simply limit yourself to the period after both of those ('73-75 for the draft, by which time the upper level post-Sputnik reforms would have hit the applicants), then factor in that freshmen went from pass/fail to grades in their second term during this period, and say MIT has statistically significant grade inflation, we have no basis for a discussion.
Yeah, you can rationalize based on outside events, but without real data on how much of the change is actually attributable to those things, you're just fitting facts into your worldview. How do we know there aren't similar events in the quarter century you've chosen that are forcing deflation?
MIT, for a bunch of reasons I've mentioned, plus intense self-selection in just deciding to apply in the first place, doesn't have this problem. Certainly didn't in the dozen years I was a member of the community, '79-'91, I never met a student who couldn't do the work.
Now, there are students who don't succeed because of financial issues (my problem, and one of at least one close friend), and personal issues (more than a few), but these are not things the admissions office can do as good a job in screening for.
Please, website owners: provide visual cues if the behavior of your website differs from normal websites.
Blocking load.sumome.com will help here.