<p>I know that the avg gpa at MIT is something like 4.2 but I'm curious what it is for engineers, particularly in eecs. At MIT's peer schools Berkeley and Stanford the average eecs gpa is 2.72 and 3.3 respectively, a pretty huge gap. I think it's pretty reasonable to expect that if it's hard to get good marks people would be pushed to work harder and learn more in the process. Heck, my friends at Stanford even concede that they might be getting more out of their educations had they gone to Berk. How does the rigor of MIT's eecs compare to these other schools. Also I've heard that MIT grades in the same way as high schools(ie. no pluses or minuses) and that's certainly a factor in the avg gpa whatever it is.</p>
<p>Actually, MIT does award plus and minus grades, but they are not posted on the official transcript, and do not factor into the GPA. The no plus/minus thing is great when you get a B-. Not so great when you get a B+.</p>
<p>
</p>
<p>I am a UC Berkeley student, as it happens, and know the engineering department relatively well, though I’m not an engineer. My own view is that while having relatively challenging grading can push some to learn, for others it’s just an unwanted source of stress. In the end, a class is a class, and you primarily want to learn stuff at as high a level as possible from it, and hopefully learn the stuff better than you would have had you studied the material yourself. For the majority of students, yes a stressful atmosphere would whip them into working properly, but for the subset of very self-motivated students, often a less stressful atmosphere can be helpful – of course, provided material is not watered down. </p>
<p>Now, whether a certain school is to hard or too easy is not for me to say, because I don’t know them all up close, but I do think in some cases Berkeley is harsher than it needs to be, and that can overfocus students on their grades and defocus them from treating a class in a somewhat more detached fashion, i.e. as just a class. I imagine overall that MIT is more similar to Berkeley in terms of grading, if you want my guess though. The end result, I think, is however that most engineers at a school like MIT will not mind so much that their average GPAs are lower than those of most other schools, as long as they’re aiming for professions in engineering and/or graduate school. Usually I think an MIT engineer will not exactly have a problem making it out there in the real world, and also not have a problem making it to graduate schools, which (for engineering especially) probably look more at things like research and recommendations than engineering GPAs.</p>
<p>MIT’s mean GPA is 4.2 out of 5. Basically, an A at MIT is worth 5 instead of 4, a B is worth 4 instead of 3, etc. The “translated” average GPA would therefore be about 3.2, slightly over a B average, which is consistent with other schools. It is also this “translated” GPA that graduate schools will request. It makes no more sense to compare the nominal value of MIT’s GPA than it does to a school that reports GPA as the average of grades out of 100. </p>
<p>I’ve never heard anyone claim that MIT isn’t rigorous and demanding! I think it’s one of a small handful of universities whose academics are renowned for being unrelenting. You could also further challenge yourself by taking graduate courses.</p>
<p>I was at Berkeley studying EECS in the early 80s and worked part time in administration in one of the specialist schools, so I had access to all grade information. I did some early analytics on the database to calculate the avg GPA of the different depts. Eng was 2.5 for lower division and 2.7 for upper division. If you made it through to upper, then it sounds like the grade policy was the same then as it is now. </p>
<p>Conventional wisdom at Cal at the time was that it was easier to get into cal than Stanford, but harder to stay in. At Stanford once you were in, you were in. </p>
<p>I was pretty depressed to see that the avg GPA in the liberal arts departments were 3.0 for lower and 3.5 for upper. Probably still the same. It always seemed like the liberal arts students were not working as hard as the engineers. They didn’t need to. The hard sciences were the same as engineering and the soft sciences were somewhere in between. </p>
<p>MIT’s average GPA is almost certainly higher than 4.2/5; see this thread <a href=“http://talk.collegeconfidential.com/massachusetts-institute-technology/1651636-average-gpa-at-mit.html”>http://talk.collegeconfidential.com/massachusetts-institute-technology/1651636-average-gpa-at-mit.html</a> where I analyze all the available data. I don’t think there is any data on difference due to department/school but my guess would be that EECS/engineering in general gave slightly harsher grades. I have no idea where the cited figure for Berkeley EECS is coming from. <a href=“Berkeleytime”>Berkeleytime; gives grade distributions for Berkeley courses which suggest the average EECS GPA at Berkeley is ~3.1. The 2.7 figure might have been accurate in the 80s but grade inflation (interpreted literally) is very real. The Stanford EECS GPA also strikes me as a little low.</p>
<p>Are you an MIT student? I don’t know what UMTYMP is–quite a long acronym.</p>
<p>The GPA may have risen in the past 15 years, but your making too many assumptions in your arguments based on incomplete data. </p>
<p>Current students or recent alumni can weigh in and say what the grading distribution is like, but I really doubt 4.4/5.0 is the average GPA Back in the 90’s, the only classes where the mean grade was the A/B borderline were the humanities classes. For engineering classes, the B/C borderline was the mean; for science classes, 60% of the class got a “B” or higher. The “A” level was reserved for roughly 15-20%. What you are suggesting is that grades have inflated a whole grade level in the past 15 years. By the way, the former president of MIT said that professors at the time felt they should be grading harder, not easier.</p>
<p>Huh, I can’t believe I haven’t googled “UMTYMP” before – apparently it’s a math program in Minnesota. </p>
<p>Professors at MIT should be grading harder, but not until they teach their classes better.</p>
<p>
</p>
<p>Well, I didn’t agree with them. I think MIT is (was) hard enough as it is. I just thought it was unlikely there was a faculty movement to make the grading easier, based on those comments.</p>
<p>I don’t know what the average GPA at MIT is, but I think the argument could be made that the quality of the MIT undergraduate student body has increased significantly in the past fifteen years.</p>
<p>A couple of points: I am a third year undergraduate at MIT and “UMTYMP” is instead a math program in Minnesota that I was in when I created my account 5 years ago. I have no idea where collegealum is getting his “data” on grade distributions. This report <a href=“Committee on the Undergraduate Program (CUP) | MIT Registrar”>Committee on the Undergraduate Program (CUP) | MIT Registrar; (page 35) gives grade distributions for AY1999-2000. Excluding freshmen whose grades at the time didn’t appear on transcripts almost half of grades given were As. Cs were barely more than 10% of grades. Page 40 of the same report gives mean GPAs for AY1999-2000 by class year. Because only decimal is listed we can’t make definite conclusions but it seems likely that the mean GPA was above 4.2 (again freshmen grades did not effect GPA). Assuming mean GPA increased by ~.15 over the past 15 years would be in line with other schools. This could be due to easier grading, better students, or better pedagogy. FSILG GPAs have also risen during this period. The 4.4 figure is obviously an estimate subject to some uncertainty but it seems much more plausible to me than FSILG GPAs steadily increasing while non-FSILG GPAs dropped which any story consistent with a 4.2 average GPA would require. </p>
<p>One alternative to grading harder would be to include +/-s on transcripts. I guess compressing grades does reduce stress though.</p>
<p>Where did I get my data? Well, I was in a major that 60% of my class was enrolled in (EECS). And I remember where the means in the classes were. In the vast majority of my classes, the mean was set such usually about 50%-60% of the class got a “B” or better. One caveat is that I tended to take core classes in other majors as extra classes to fulfill my distribution requirement rather than upper level classes; upper level classes sometimes are graded easier. In no technical class that I was in was it remotely close to 90% of the people getting A’s or B’s. I took probably 10 core classes outside of my major (typically sophomore level core classes.) </p>
<p>You are looking at page 40 and it is apparent from this data that the GPA is around 4.2.</p>
<p>Freshman-- 3.8, 3.8 (pass/fail, not counted)
Soph – 4.1, 4.2
Junior – 4.2, 4.3
Senior – 4.3, 4.3</p>
<p>These GPAs are “by term,” which I take to mean they are not cumulative GPAs, but the GPA for classes taken that term. I would expect the cumulative GPA to be (4.1+4.2+4.2+4.3+4.3+4.3)/6 = 4.25 based on this. These grades are partly brought up by the HASS classes, where the average GPA appears to be ~4.4. So given that there are 8 HASS classes required, this is like a year of classes. So the technical GPA will be under 4.2. Add to that that the engineering classes in my observation were graded harder than the science classes, so the average engineering GPA may have been about 4.0/5.0. </p>
<p>
</p>
<p>Well, the first rule of statistics is that you need a random sampling in order to determine what is going on in a larger population. Though it is a large sampling, the FSILG subpopulation is decidedly not a random sampling of the undergrad population. If the FSILG gpa statistics can be trusted and are not self-reported, the average GPA rise of .15 pts would imply at least a rise of 0.04 in the undergrad GPA. There are more scenarios out there that could explain why there might be a GPA rise in the FSILG and not in the general population. Maybe the massive changes in the FSILG system (all freshman on campus) meant there was less hazing of freshmen, and they were better prepared to do well in upper level classes. Maybe the elimination of second semester pass/fail impacted the cumulative GPA. Most people think these classes are easier than the upper level classes. </p>
<p>As for the GPA change at large, it may not imply anything about a particular major or even technical majors in general. The OP asked about engineering (though they’re probably in graduate school now). The Institute has seen the percentage of management majors go from nothing to a pretty significant minority of the campus. I’d expect management to be graded more like a HASS. This factor alone could account for a GPA shift, and it doesn’t tell you anything about what you might expect as an engineering major. </p>
<p>I will agree that upper level classes are graded more leniently than core classes. I would also suspect that classes which release detailed data on grade distributions have lower GPAs than average largely because both are common in large core classes. I also looked at enrollment statistics which only go back to AY1998-99. At no point was EECS close to 60% of undergrads. It is also not the case that management is now a significant minority of the campus. Enrollment in course 15 peaked over a decade ago at close to 10% of students and has since plummeted in popularity with just 12 majors (~1%) in the class of 2017. I agree it is apparent that the mean GPA in AY1999-2000 was around 4.2 but it seems unlikely given national data it has not increased since then.</p>
<p>It is true that the OP asked about engineering majors although they cited 4.2 as the average MIT GPA. I would also guess that the average engineering GPA is less than the average GPA which I estimate at 4.4. The mean reason to focus on estimating average GPA instead of engineering GPA is that there is data that allows reasonable estimates of the former while I’m unaware of any data on the latter.</p>
<p>
</p>
<p>Well, if you are referring to college GPAs across the country by “national data,” I don’t think national trends necessarily reflect what is going on at MIT. </p>
<p>For virtually every college on which there is historical GPA data there has been an increase in GPA over time. The few exceptions are generally schools that has instituted strict policies to combat grade inflation. It’s possible MIT is an outlier although given what data we have it seems unlikely.</p>
<p>I don’t have a cite and don’t feel like digging one up, but 4.2 has been published as the average GPA in the Tech in recent years. The Tech isn’t always right, but I’m willing to trust it on this.</p>
<p>Let me address the continued confusion over the 4.2 number which has is unfortunately widely believed. There are three main problems with the citation (here is the article you are thinking of <a href=“http://tech.mit.edu/V128/N5/cupreport.html”>http://tech.mit.edu/V128/N5/cupreport.html</a>)
(1) The 4.2 average is for sophomore only. The most recent data we have on GPA by class year is unfortunately from the 1990s (<a href=“Committee on the Undergraduate Program (CUP) | MIT Registrar”>Committee on the Undergraduate Program (CUP) | MIT Registrar) when sophomore GPAs were close to .1 less than average GPAs (see pages 35 and 40 of that report).
(2) The 4.2 number isn’t fully supported by the report which the Tech article cites <a href=“https://web.mit.edu/committees/cup/mit-only/P-NRandExp.pdf”>https://web.mit.edu/committees/cup/mit-only/P-NRandExp.pdf</a>. Somewhat confusingly the average GPA for sophomores is both given on page 8 and on page 15. The report notes on page 15 that there is a small discrepancy between the data sources but views it as insignificant. This seems unlikely as the data on page 8 gives 4.3 as the typical average sophomore GPA while page 15 gives 4.2. I have not figured out which number is actually correct or why they are different. In any case these numbers are higher than the average sophomore GPA from the late 1990s supporting the thesis that MIT sees grade inflation (as with almost every other college).
(3) The data is 7+ years old at this point. Both the previously mentioned data as well as data from FSILGs supports the idea that MIT sees grade inflation and hence a 7+ year old estimate is likely to be an underestimate of the order of .05-.1.</p>