Ivy League Grading Rubric leaked

Pffffft I graduated Harvard with a C average, it can be done, but you have to be awesome like me.
 
The link took me to an ad to buy NYT, which made me throw up in my mouth, just a little.
 
Someone I know in Harvard Engineering says the classes are all curved to a B+/A-. Here at Michigan, they're all curved to a B-. Why? Beats me.
 
jpower said:
Someone I know in Harvard Engineering says the classes are all curved to a B+/A-. Here at Michigan, they're all curved to a B-. Why? Beats me.
Click to expand...

Too many college kids think they deserve the grade without having to work for it?
 
jpower said:
Someone I know in Harvard Engineering says the classes are all curved to a B+/A-. Here at Michigan, they're all curved to a B-. Why? Beats me.
Click to expand...

Private schooling vs public schooling. Gotta keep the money rolling in. That won't happen if you fail the kids out.
 
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended.
 
The real engineers are down river a bit. There's always the joke as to how the Harvard Bridge got its name.
 
docmirror said:
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended.
Click to expand...

See that's reasonable. I had a professor last semester who curved me down. I got a 90% on the first exam after screwing up one inverse Laplace transform--the final step of the problem--and getting no partial credit for it. Normally a 90% would be a good grade...great even at this place. But 40% of the class got a 90% or above, which, with the curve, put me in the B/B- range for that exam.
 
jpower said:
See that's reasonable. I had a professor last semester who curved me down. I got a 90% on the first exam after screwing up one inverse Laplace transform--the final step of the problem--and getting no partial credit for it. Normally a 90% would be a good grade...great even at this place. But 40% of the class got a 90% or above, which, with the curving down, put me in the B/B- range for that exam.
Click to expand...

Yup, that is very unfair. And I did give partial credit for sequence process calcs. If someone were doing a linear and they got the matrix setup right and maybe inverted, or missed a sign somewhere, I would give them the benefit of half points or so. Same with integration. And I never used a curve for grading in the class. Who knows, one semester you could get a bunch of Feynman's and the next a bunch of Rain mans. (no pun intended, but fun anyway)
 
docmirror said:
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended.
Click to expand...

Sounds like my approach when I was an adjunct instructor of engineering. I had a guy come to me and tell a sob story about needing a B to make a GPA to remian on track. My reply was you will get the grade you earn, anything from A to F is possible.

BTW, I graduated with a 2.8, IIRC. When I went back 7 years later to grad school, as far as I could tell, my similar work resulted in a 3.5, both out of 4. Based on observation, it would now likely be a 4.0.

Cheers
 
Last edited:
I had a professor that announced on the first day of class he graded on the curve. Everyone was nodding heads, like "oh yeah, sweet" and he then continued on to say the Bell Curve. At the end of the semester the grades given would be one A, two Bs, two Ds, one F, and the rest would be Cs, and that's the way all papers submitted would be graded. My first paper I turned in was a 39/100. I went to the professor and asked what was up with that? He said that was in the top half of the class, and not that far from being a B. I said screw that, I looked around at the next class, determined which three would be the Ds and F, and gave minimal effort the rest of the semester and took my C.
 
Algebra and thermodynamics has been around for a long, long time. I never altered my curriculum for more than a decade, and each year it was the same stuff. About half would get it, about 20% would never get it, and the other 30% could have taught it with a little prep. Engineering students are usually highly motivated, and internally driven, so they like knowing where they stand. I wasn't on staff either, and I can count on one hand the number of times I had to adjust a grade after the fact. But - I did have to do it when a student presented a compelling case. Sob stories lasted about 8 seconds, just like a 'bull' ride. lol
 
docmirror said:
Yup, that is very unfair. And I did give partial credit for sequence process calcs. If someone were doing a linear and they got the matrix setup right and maybe inverted, or missed a sign somewhere, I would give them the benefit of half points or so. Same with integration. And I never used a curve for grading in the class. Who knows, one semester you could get a bunch of Feynman's and the next a bunch of Rain mans. (no pun intended, but fun anyway)
Click to expand...

Interesting approach....:confused:

In your first post you said

"
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended "

I am just a poor ol commercial contractor building multi million dollar, very complex structures.... If I accidently leave out a critical beam /column and the building collapses... Do I get credit for the part that still stays standing..:nonod:.

My point is, either the task,( for me) or the test ( for the student) conclusion is right or wrong....

In the real life, you don't get credit for partially successful outcomes... Why give students that option ? :dunno:

Thanks in advance for your answer sir... ;)
 
Here's an app that determines what your grade at Harvard would be based on a series of complex algorithms.
http://www.whatwouldmyharvardgrade.be/

The thing that I hate about schools doing this is that I go to a school where your grades are actually earned. My first semester in college, my math prof. had just moved to CMU from Harvard, and he said that the equivalent of an A+, A, and A- in his class at Harvard was equivalent to an A, B, C at Carnegie. (we actually don't even grade on a scale either. An A is a 4.0, B is a 3.0, etc. with nothing in between, which means lower grades need to be given in order to distinguish performance).
Unfortunately, many recruiters won't necessarily understand how our school does things, see a 3.7 at my school and compare it on the same scale with someone with a 3.9 from someone at the Ivys. Ultimately, it doesn't really matter how it's done as long as it's uniform.
 
If you ever see a pile of Harvard applications, the high school GPAs run around 4.6 out of a 4.0:eek: It ain't just the Ivies, it is the Ivy feeder schools as well.:rofl:
While at state schools the Asians and Indians cheat on everything, and the minorities are socially progressed. All grades are make believe(except possibly some engineering classes.) Besides I've never met a smart person that brags about grades, grade talk is all SWPL posturing nonsense.
 
docmirror said:
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended.
Click to expand...

I had a professor like that. The first day of class everyone hated him, but by the second week he was pretty cool. His motto was "I don't care about your GPA and you shouldn't either. It's just a number and this class won't reflect anything about you as a person so don't use your GPA as a crutch". A lot of the straight A students hated him because you had to work pretty dang hard to get an A (I got a B). "Professor you RUINED my GPA!!!" "I didn't ruin anything, you ruined it."
 
ClimbnSink said:
If you ever see a pile of Harvard applications, the high school GPAs run around 4.6 out of a 4.0:eek: It ain't just the Ivies, it is the Ivy feeder schools as well.:rofl:
While at state schools the Asians and Indians cheat on everything, and the minorities are socially progressed. All grades are make believe(except possibly some engineering classes.) Besides I've never met a smart person that brags about grades, grade talk is all SWPL posturing nonsense.
Click to expand...

It is very easy to get above a 4.0 in high school. Honors classes are weighted differently from a regular class. (on a 4.5 scale I believe) and AP/College classes are weighted on a 5.0 scale.

In a honors class an A is a 4.5, a B is a 3.5, a C is a 2.5, a D is a 1.5, and an F is a 0.5.

So an A in an AP/College class is a 5, a B is a 4, a C is a 3, and D is a 2, an F is 1.
 
N801BH said:
Interesting approach....:confused:

In your first post you said

"
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended "

I am just a poor ol commercial contractor building multi million dollar, very complex structures.... If I accidently leave out a critical beam /column and the building collapses... Do I get credit for the part that still stays standing..:nonod:.

My point is, either the task,( for me) or the test ( for the student) conclusion is right or wrong....

In the real life, you don't get credit for partially successful outcomes... Why give students that option ? :dunno:

Thanks in advance for your answer sir... ;)
Click to expand...

So, Mr. Contractor...if you build a multi-million dollar facility, and accidentally install a door knob without a lock on one door instead of a locking door, your client is free to withhold 100% of their payment to you?
 
N801BH said:
Interesting approach....:confused:

In your first post you said

"
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended "

I am just a poor ol commercial contractor building multi million dollar, very complex structures.... If I accidently leave out a critical beam /column and the building collapses... Do I get credit for the part that still stays standing..:nonod:.

My point is, either the task,( for me) or the test ( for the student) conclusion is right or wrong....

In the real life, you don't get credit for partially successful outcomes... Why give students that option ? :dunno:

Thanks in advance for your answer sir... ;)
Click to expand...

You get partial credit if you build 8 of them in 50 minutes.

I took many DE tests that consisted of one problem, you'd struggle to finish it in the time alloyed. Some were unsolvable. Just stating the problem was unsolvable would get you 0 points.
 
Last edited:
JeffDG said:
So, Mr. Contractor...if you build a multi-million dollar facility, and accidentally install a door knob without a lock on one door instead of a locking door, your client is free to withhold 100% of their payment to you?
Click to expand...

Depending how the contract is written, yes. Of course the difference is he can go back and install the lock on the door when notified. The student isn't allowed to go back and fix a mistake. So the comparison really isn't valid.
 
N801BH said:
Interesting approach....:confused:

In your first post you said

"
Here's the spiel I used to give when I was teaching math or eng at the start of the class: 'I get paid the same if you all get As or Fs, or any combo in between. the school will still cash your tuition check no matter if you pass, or fail. There is no grading on a curve, there are no fancy adjustments, or interpolations of the raw data. You will be presented with the material, it will be reinforced by homework, you will be tested, and your grade will accurately reflect your retention of the material over the semester. Please bring your drop class slip to me for signing.' Most of my classes were very well attended "

I am just a poor ol commercial contractor building multi million dollar, very complex structures.... If I accidently leave out a critical beam /column and the building collapses... Do I get credit for the part that still stays standing..:nonod:.

My point is, either the task,( for me) or the test ( for the student) conclusion is right or wrong....

In the real life, you don't get credit for partially successful outcomes... Why give students that option ? :dunno:

Thanks in advance for your answer sir... ;)
Click to expand...
In real life constant mistakes are made and partial credit is almost always given.
 
EdFred said:
Depending how the contract is written, yes. Of course the difference is he can go back and install the lock on the door when notified. The student isn't allowed to go back and fix a mistake. So the comparison really isn't valid.
Click to expand...

Neither is his assertion that "In the real life, you don't get credit for partially successful outcomes"

There are very few real-life situations where the result is either perfection or nothing. Real life is all about partially successful outcomes.
 
When I was going through undergraduate, the grading curve was pretty much the standard thing and I was more or less a B student. When I went in to grad school, the whole curve shifted back a letter grade and I was a C student. It was that way for pretty much everyone - the former A students were now B students, and the former C students got cut a little more slack so they didn't go in to D territory and fail.
 
jesse said:
In real life constant mistakes are made and partial credit is almost always given.
Click to expand...

Not really. We either get the sale or we don't. There is no "oh, you made a great sales pitch, and we are going with company X, but here's some money anyway because your sales pitch was so good."

And in my personal life, there is absolutely no partial credit given. It is most definitely a binary switch.
 
EdFred said:
Not really. We either get the sale or we don't. There is no "oh, you made a great sales pitch, and we are going with company X, but here's some money anyway because your sales pitch was so good."

And in my personal life, there is absolutely no partial credit given. It is most definitely a binary switch.
Click to expand...

Sweet. I want to hire you. First mistake you make I can claw back all the salary I have ever paid to you, because, you know, no partial credit given at all, right?
 
The problem with grading on a 90% standard is that you nail people for trivial mistakes; it becomes a test of precision and memorization. If that's what you want to test, fine. Otherwise, it's not a useful exercise.

The approach we took in an undergraduate physics curriculum was to make the tests HARD and generally open book, and then grade on a curve. Partial credit was given liberally where appropriate (but NOT if it indicated a fundamental misunderstanding -- simple arithmetical errors got partial credit, but applying energy conservation to a collision problem didn't). We looked for a bimodal distribution, as there was generally a small number of students who just blew it completely, generally by never showing up to lectures and never turning in homework. If there was no second mode, there were no Ds.

Quite a number of the comments in this thread indicate a misunderstanding of what "the curve" is or what its purpose is. The 90% standard if fundamentally arbitrary. And a test that conforms to that is not a good test.

There were a fair number of students who didn't like getting 70% on an exam and getting an A for it, believe it or not, and especially getting 50% and getting a B. We generally designed tests so that the median would fall around 60%, and assigned a B there. That way, the test actually tests whether people understand or not, not whether they can carry a minus sign through a complex calculation. Minus signs can matter tremendously, but focusing on that in an intro course misses the point; I'd rather have the students thinking about concepts. This DOESN'T mean you can't have a class where everyone gets an A. If I design a test to a 60% median, and the class as a whole aces the test, giving everyone an A (except for the test designer) is reasonable. This happened a couple of times. It's all about exceeding expectations. As is life.

I question any central "guidance" on this sort of thing. Even talking about that indicates a poor understanding of how universities work. They do NOT work like corporations, despite media focus on university leadership. More like anarchy. Tenured faculty have all the power.

People seem to want grades to be objective. This seems to be based on a fundamental fallacy that an exam is an absolute. It almost never is. They are designed by humans to test specific things, and as such, are always subjective to some extent.
 
Last edited:
Good thing my Uni didn't grade on a curve, the Malaysian kids would have had to hire bodyguards :)
 
JeffDG said:
Sweet. I want to hire you. First mistake you make I can claw back all the salary I have ever paid to you, because, you know, no partial credit given at all, right?
Click to expand...

If our company makes 0 sales, there is $0 to pay out. There is no partial credit. As I said, it's binary.
 
EdFred said:
If our company makes 0 sales, there is $0 to pay out. There is no partial credit. As I said, it's binary.
Click to expand...

But you said NO partial credit, so if you make the sale and someone makes a minor mistake, you get paid zero.
 
JeffDG said:
But you said NO partial credit, so if you make the sale and someone makes a minor mistake, you get paid zero.
Click to expand...

What minor mistake? The sale is either made, or it isn't. There isn't a third option.
 
EdFred said:
What minor mistake? The sale is either made, or it isn't. There isn't a third option.
Click to expand...

Post sale.

Let's say you sell someone 100,000 widgets. 99,999 are perfect, but one is defective. By the "no partial credit" philosophy, you get paid nothing for any of them.
 
N801BH said:
Interesting approach....:confused:

In your first post you said

I am just a poor ol commercial contractor building multi million dollar, very complex structures.... If I accidently leave out a critical beam /column and the building collapses... Do I get credit for the part that still stays standing..:nonod:.

My point is, either the task,( for me) or the test ( for the student) conclusion is right or wrong....

In the real life, you don't get credit for partially successful outcomes... Why give students that option ? :dunno:

Thanks in advance for your answer sir... ;)
Click to expand...

Well, life isn't perfect, and neither is the engineering that goes into civil, elec, mech, chem, phys systems. Sometimes people miss stuff and a bridge comes down. http://scienceprogress.org/2008/01/design-flaw-likely-cause-of-minneapolis-bridge-collapse/

That bridge was put up in 1964, and it came down in 2007 in dramatic way. The engineering co that put it up got 43 years out of it. Sadly, due to some construction materials, and a problem with some of the gussets it did not fulfill it's lifetime guarantee that it would never fail. Oh wait, the didn't offer a lifetime guarantee, they did the best they could and the firetrucks, and ambulances, and millions of plain old traffic miles might have saved thousands of lives during it's life.

I'm a pretty black and white kind of guy, but I know if you are going to build a bridge, there will be two or more people checking your work, and finding flaws. In software we call them bug reports. In electronic hardware they are flaw, etc. That's real life, and sometimes perfection cannot be achieved. If any of my students were perfect all the time, it would be a boring world. So, partial credit is certainly warranted if it's in my judgement that the process was correct, but an arithmetic, or sign error produced the wrong absolute result.

Yes, I've been guilty of making a mistake on the job, which cost a company some money in the past. I made good on it, and our company moved forward and we all learned that mistake so we don't make it again.
 
JeffDG said:
Post sale.

Let's say you sell someone 100,000 widgets. 99,999 are perfect, but one is defective. By the "no partial credit" philosophy, you get paid nothing for any of them.
Click to expand...

No, we replace the 1, still giving them 100,000 perfect widgets. As I said, there is no third option.
 
EdFred said:
No, we replace the 1, still giving them 100,000 perfect widgets. As I said, there is no third option.
Click to expand...

So, you're claiming partial credit with the opportunity to correct. You did not perform perfectly, you should receive nothing.
 
JeffDG said:
So, you're claiming partial credit with the opportunity to correct. You did not perform perfectly, you should receive nothing.
Click to expand...

I am not claiming that at all, you are baiting me to do so. It won't work. The order is for 100,000 working units they get 100,000 working units. Not 99,999, 100,000. Why is that a difficult concept to grasp?
 
EdFred said:
I am not claiming that at all, you are baiting me to do so. It won't work. The order is for 100,000 working units they get 100,000 working units. Not 99,999, 100,000. Why is that a difficult concept to grasp?
Click to expand...

You delivered a defective order, and you then insist that the customer let you correct it.

If you say that there is absolutely no partial credit, you shouldn't be paid anything. That seems to be the standard if demanding that an exam question not receive any partial credit, unless you're willing to let the student go back after the exam is marked and fix stuff.

When you delivered the widgets, you handed your exam in. The customer marked the order and found that it was not correct. No partial credit means a zero score.
 
JeffDG said:
You delivered a defective order, and you then insist that the customer let you correct it.

If you say that there is absolutely no partial credit, you shouldn't be paid anything. That seems to be the standard if demanding that an exam question not receive any partial credit, unless you're willing to let the student go back after the exam is marked and fix stuff.

When you delivered the widgets, you handed your exam in. The customer marked the order and found that it was not correct. No partial credit means a zero score.
Click to expand...

They don't pay until the order is complete. So, as I said before, no partial credit.

And that's not even getting into the personal side of my life.

Neither of which is an accurate comparison to the exam.
 
Last edited:
JeffDG said:
You delivered a defective order, and you then insist that the customer let you correct it.

If you say that there is absolutely no partial credit, you shouldn't be paid anything. That seems to be the standard if demanding that an exam question not receive any partial credit, unless you're willing to let the student go back after the exam is marked and fix stuff.

When you delivered the widgets, you handed your exam in. The customer marked the order and found that it was not correct. No partial credit means a zero score.
Click to expand...

Actually it would mean he got a 99,999/100,000 which is roughly a 99 A.
 
You must log in or register to reply here.
Back
Top