U of A eliminates bell curved grading

[KOgear]

Originally Posted by Phanuthier

In an interview?

In one of my interviews, I was asked "If you were to put a donut in a microwave, and microwave it for a long period of time (i.e. 30 minutes), what shape would it take on?"

PS : I'm going to use that one if I get a chance to interview a new grad sucker

No this is how crazy the tests were, another one was how much heat would be lost in x amount of time if a portion of the airplane cabin ripped off mid flight.

I would just write down a bunch of equations and derive them in the hopes of getting part marks.

Please spare them, or if you do ask a hard one at least make it realistic.

[J epworth]

UWO never did bell curving for faculty of science, and I believe it gave a better representation of how much you know in the class. Organic chem the class average was 61 percent, and your mark was your mark. I felt other schools that bell curve courses like that unfairly made students who would have got 70 percent in that course look like they aced the course, when in fact they didn't know 30 percent of the material properly

[MarchHare]

Some of the questions were ridicolous like, how long would it take to cool a cup of coffee in a vacuum. When would you ever need to know this!!

I have absolutely no idea how to solve that problem now, but it seems like the type of question that would have been relatively easy when I was taking 1st year chemistry. I guess you'd use basic thermodynamics or Newton's Law of Cooling or something else that I no longer remember from my undergraduate days (Boyle's Law is popping into my head, but I think that applies to gases, not liquids). I assume you were supplied with the the volume of the coffee and the surface area of the mug, as those would be necessary to calculate the cooling time (cooling will be faster with smaller volumes and greater surface areas and slower for the opposite).

As for "when you would ever need to know this": anyone who has that attitude (and MANY people do), is missing the entire point of a university education. If you only want practical job training, go to a technical/vocational school like SAIT. Universities exist to provide a well-rounded education, not to act as workplace training facilities.

[Edit]

No this is how crazy the tests were, another one was how much heat would be lost in x amount of time if a portion of the airplane cabin ripped off mid flight.

What's so crazy about that question?

[d_phaneuf]

Originally Posted by wpgflamesfan

What happens when you have the same class being taught by two different teachers with one giving super easy exams and the other giving much harder exams?

That's the beauty of the curve. It eliminates any of these small biases and compares people relatively rather than objectively, and at the end of the day that's closer to what you'll see in real life.

People are going to go to ratemyprofessor type sites and likely avoid the harder exams one

[SebC]

Originally Posted by KOgear

Some of the questions were ridicolous like, how long would it take to cool a cup of coffee in a vacuum. When would you ever need to know this!!

It's in a vacuum to make your life easier. Disregard convection, disregard conduction, radiation only. But yeah, my exams at U of A had class averages ranging from ~40 to ~94, so if they're not curved, they sure as hell better be scaled. Profs had leeway in the exact class average anyways (e.g. B, B+) so they could reward better years and punish worse classes even on the old system.

[bigalspratty]

Sounds like as a whole they are not removing the bell curve. Teachers will retain the right to curve a class based on specific criteria. It will be their judgement call and with approval from the faculty dean and whoever else.

[Bobblehead]

The article has been updated:

The university’s decision to abolish the requirement of the bell curve was finalized last week, and will be implemented in the fall semester across all faculties and departments on campus

So it sounds like up to this point ALL of UofA has been curved.

[chemgear]

Originally Posted by Bobblehead

The article has been updated:

Updated the OP for the . . . update.

[Kipperriffic]

nvm

[bigalspratty]

Originally Posted by Bobblehead

The article has been updated:


So it sounds like up to this point ALL of UofA has been curved.

I think it was class dependent. I took classes of 20 people and the instructor said it wasnt curved. The Phys-Ed faculty made it an option depending on the size of the class.

[SebC]

It might be hard to believe, but the Faculties at the U of A don't always follow the rules set out by the University. For example, the university rules state that exams must take place at the same time/day as the class the exam is for, which the Faculty of Engineering simply ingores.

[Lt.Spears]

Originally Posted by MarchHare

No, that's not what marking on a curve is; that's adjusting the percentage required to achieve a certain letter grade.

Using a bell curve to assign grades means that class rank is statistically modeled using normal distribution and letter grades are assigned accordingly based on how you perform on an exam in comparison to your peers, not based on your absolute test score.

http://en.wikipedia.org/wiki/Bell_curve_grading

What you proposed is still basically curving, if the average is a 30% and it takes a 50% to get an A, then thats on a curve...

Unless you are stating that a professor tells his students before hand that 50% is an A, in which case.... yes, that isn't really on a curve because he does not know the statistics of the grades yet, but that is moot because it probably would never happen (and if it were, would be based on historicals, which in essence is grading on a curve)

[MarchHare]

Originally Posted by Lt.Spears

What you proposed is still basically curving, if the average is a 30% and it takes a 50% to get an A, then thats on a curve...

Unless you are stating that a professor tells his students before hand that 50% is an A, in which case.... yes, that isn't really on a curve because he does not know the statistics of the grades yet, but that is moot because it probably would never happen (and if it were, would be based on historicals, which in essence is grading on a curve)

What you're describing is scaling, not grading on a curve. If it's possible for every student in the class to get an A (or an F), then you're not using a curve.

Bell curve grading is when a certain percentage of the class will get an A, B, C, etc., and this is stated in advance. Usually, but not always, this is designed to match the statistical normal distribution (bell curve), hence the name.

Say, for example, you have a class where the prof is grading on a curve. He might say that the top 15% will get an A, the next 30% will get a B, the next 30% will get a C, the next 15% will get a D, and the bottom 10% will fail. Your grade is determined entirely by how well you performed in relation to your peers, not at all based on your absolute test score.

[littlereddevil]

Originally Posted by diane_phaneuf

People are going to go to ratemyprofessor type sites and likely avoid the harder exams one

I find that site terribly useless and misleading because I've had some terrible profs and they've gotten very good reviews on that site.

To each to his own I guess...

[GGG]

Originally Posted by KOgear

Some of the questions were ridicolous like, how long would it take to cool a cup of coffee in a vacuum. When would you ever need to know this!!

I loved questions like this because they are actually really simple but in order to answer the question you need to really understand the subject matter. I don't remember the solution to the question above but its basis is around Black Body radiation and planks law. Since there is no convection or conduction in a vaccuum the only way to lose energy / heat is to radiate. So if you know the temperature of the cup of coffee you can determine its rate of energy release. If you assume that its rate of radiation is linear (true over a small delta T) then you simply calculate the amount of energy needed to be emitted to lower the temperature of the cup of coffee.

So if you understand Black body radiation concepts the question only takes 1 or 2 minutes of calculating. If you don't you start trying to consider convection and conduction and waste time doing complex calculations.

I find these types of questions separate true understanding from knowing how to apply a formula. They are also the most frusterating to see the correct answer on because you skip the question, get no marks for it and the solution is 4 or 5 lines of calculations compared to the rest of the test requiring pages to solve each question.

To discuss the original topic I think that curving marks make sense. (ie if everyone fails you bump up all the grades) but applying a bell curve to classes shouldn't be done outisde of the 400 person psych 101 classes and in theory classes of that size should curve themselves. My defintion being that a curve is any change of grading applied where as the bell curve represents changing the marks so that it fits a Normal distribution.

[SebC]

Originally Posted by GGG

I loved questions like this because they are actually really simple but in order to answer the question you need to really understand the subject matter. I don't remember the solution to the question above but its basis is around Black Body radiation and planks law. Since there is no convection or conduction in a vaccuum the only way to lose energy / heat is to radiate. So if you know the temperature of the cup of coffee you can determine its rate of energy release. If you assume that its rate of radiation is linear (true over a small delta T) then you simply calculate the amount of energy needed to be emitted to lower the temperature of the cup of coffee.

So if you understand Black body radiation concepts the question only takes 1 or 2 minutes of calculating. If you don't you start trying to consider convection and conduction and waste time doing complex calculations.

Or you could use rate of radiation varies with T^4, like you're supposed to (since calculus is a pre-requisite to heat transfer).

[GGG]

Originally Posted by SebC

Or you could use rate of radiation varies with T^4, like you're supposed to (since calculus is a pre-requisite to heat transfer).

I dont like integrals....and really T^4 is an approximation as well just a better one.

[ernie]

Most experienced profs can write an exam that will generate a bell curve or darn close. This doesn't do much. Nor does this change grades from being scaled...they just can't be forced into a normal distribution.

It's not a big deal.

[Jake]

I could be incorrect, but the way I am reading this is that a bell curve could still be applied based on the professors judgement.

I don't really care though. I never had an exam bell curved at U of C, despite a few exams having class averages in the 50s.

[Yeah_Baby]

U of C is clearly the education mecca in the province of Alberta. That's what I'm getting out of this thread.