/sci/ - Science & Math

[Anonymous No.9077825]

Does /sci/ like Bellcurve grading?

[Anonymous No.9077831]

>>9077825
No, college classes for the most part aren't competitive. They're mills.

[Anonymous No.9077837]

The professors who do this are the same kind who take personal offense when you don't show up to class.

[Anonymous No.9077862]

Tfw the average grade in engineering design module was 20%
TFW to rectify this they added 30% onto anyone who didn't get 0

>ended up with 75% despite almost failing the exam.

Kek

[Anonymous No.9077869]

>>9077825
Nah. My favorite courses in college were the ones where the professor took everything into consideration from the semester and handed you a grade at the end, regardless of average.

[Anonymous No.9077893]

>>9077869
Yeah, the last math course I took in undergrad was a topics course in differential geometry:
>HW was ~10 problems long, you were asked to do around half
>You had an hour long session every week with the professor where you orally presented your solutions and talked about any problems you had in the class one-on-one
>Final exam was in the exact same format but you had 24 hours instead of a week to do the problem set and the problems were much, much easier.
>Super relaxing final because you were just doing something that you had been doing every week with a guy that you knew pretty well.
>Grade based on how familiar he thought you were with the material and whether or not you went in a reasonable first direction with your proof, rather than whether or not you completed the problems.
Feels bad that this is apparently just how higher education works in the Netherlands. If only I had known that when I was 18.

[Anonymous No.9077909]

>>9077825
No. Grades should not be based on average marks. they should be based off of the performance of an individual on an exam, not as a collective average

[Anonymous No.9077968]

Protip: standard grading is also grading to a curve

[Anonymous No.9077989]

>>9077909
Why? Statistically that's pretty much the worst way to test knowledge of a subject.

[Anonymous No.9078035]

curve grading where you peg the highest and lowest grades as As and Fs is garbage

curve grading where you determine that typical mastery of the material should get you a course average of 50 and set the thresholds for As and Fs at 80 and 30 is fine

[Anonymous No.9078227]

>>9077825
Yes anyone who doesn't like it is a brainlet who is mad they got a bad grade

[Anonymous No.9078352]

>>9077869
My advisor used this method for most of his upper-division courses. It was stressful as fuck sometimes, but it almost always worked out in my favor.

[Anonymous No.9078386]

>>9077825
No if everyone in the class knows the material required to pass the class and get a degree/certified they should pass

>but muh arbitrary benchmark to make people fail so I can feel better about myself

[Anonymous No.9078399]

I really like it because I work much harder than most of the other students. sadly most of my professors don't use it

[Anonymous No.9078411]

>when you end calc II with a 122

[Anonymous No.9078419]

>>9077825
Grades should curve to a B

[Anonymous No.9078424]

>>9077869
> test scores dont matter just do your best

[Anonymous No.9078429]

>>9077825
average should be a 50% if you want it to be symmetric

[Anonymous No.9078431]

>>9077837
But those professors are the best.

[Anonymous No.9078436]

>>9077825
>At MIT, Berkeley, Stanford, Harvard
>Everyone there is going to contribute well to society
>Let's just give them all Cs so cletus from University of Bumfuck Nebrahoma can get into grad school but they can't

[Anonymous No.9078437]

>>9078429
the average should be targeted to 30 or so. nobody cares about the differences between failing terribly and failling horrendously, but we do care about the differences between doing well and doing amazing in a course. if the average is 30, that gives us the majority of the scale to separate differences in the most intelligent

[Anonymous No.9078456]

Bell curve grading just means you go to a low tier university with a hard ass professor.

Ivys all have grade inflation because administrations just got tired of privileged students whining.

[Anonymous No.9078463]

>>9078456
>career already fucked because low tier uni
>professor has an ego complex and ass fucks me even by bell curve grading
>students at top tier unis don't have to deal with this shit because their professors are well paid and have fruitful research

why is this even a thing

[Anonymous No.9078464]

>>9078456
grade inflation is the result of poorly applied grading curves

the standard 70/80/90-C/B/A system is also a grading curve

there's nothing about grading curves that inherently leads to grade inflation

[Anonymous No.9078467]

I like bell curves so long as the professor doesn't average down. For example, if the average on a test is 85%, then rounding that down to a C instead of letting it be a proper B is frustrating as fuck since you essentially have no margins of error.

Best grading scale is a bell curve with the caveat that if you score a <90% raw score in the class then you get an automatic A, regardless of class average.

[Anonymous No.9078475]

>>9078467
>get a 50
>automatic A

[Anonymous No.9078485]

Sounds autistic and unfair. Can someone explain the reason behind it?

[Anonymous No.9078491]

>>9078485
Some professors have this notion that "their class is only for serious students", so to them, a 100% pass rate would mean they are doing their job wrong.

[Anonymous No.9078504]

>>9078485
The teacher makes a judgement call about what the minimum level of mastery they expect from students is. That level of performance is then assigned a D or a C (depending on your institution's policies). They then assume that grades will tend to follow something somewhat like a normal distribution with standard deviation of about ten points, so as a shortcut they set the boundaries between grade bins as roughly standard deviation marks away from the mean.

The most common implementation is that the minimal level of mastery of material tends to be associated with an overall score of 70-75 percent in a course. A B is then 80-85, A is 90-95, and D is 60-65. In this model, a grade of A represents an enormously high score relative to the mean - you scored about two standard deviations higher than the rest of the class.

However, sometimes a professor decides that their material is harder than normal, and minimal level of mastery is actually associated with a different grade. They might decide that minimal mastery is associated with an overall score of 50, or 80. The bins are adjusted accordingly.

[Anonymous No.9078509]

>>9077862
wow it's almost as engineers aren't actually that smart

[Anonymous No.9078512]

>>9077862
wow it's almost as if engineers aren't actually that smart

[Anonymous No.9078523]

>>9077825
>be genius in a class of brainlets
>100sigma up from the average
>get one grade above the 2sigma guy with a B

[Anonymous No.9078527]

>>9078475
Sorry, I meant >90%, mistyped.

[Anonymous No.9078529]

>>9078527
Sorry I meant to not post here ever again.

[Anonymous No.9078534]

>>9078485
It's an easy way to structure a class. You make the class as hard as you please, then fail the lowest 20% of the class and give an A to the highest 20% of the class. This is somewhat fine, but it's dependent on the assumption that high test scores indicate mastery, which some people claim that they do not.

[Anonymous No.9078541]

>>9078529
Rude

[Anonymous No.9078552]

>>9078534
they indicate a combination of effort and mastery. either alone shouldn't get you an A, and if you have mastery without effort why are you even bothering with college in the first place

[Anonymous No.9078554]

>>9078485
Let's say you have a bunch of material to cover but you're a shitty teacher and half the class would fail to learn. Instead of adding new course materials (bookwork, etc.), just give insanely hard tests and curve it to where a 50 is an 80.

[Anonymous No.9078568]

>>9078552
I agree with you, but sometimes I get an A in a class and then 6 months later I can't remember much about what we learned

[Anonymous No.9078574]

>>9078568
Everyone forgets things if they don't keep practicing the material, I don't think that really says anything about the course.

[Anonymous No.9078585]

>>9077825
bell curve grading is the most retarded thing a teacher could do

[Anonymous No.9078588]

>>9077825
If you get a 90% on the every test you should get a 90% in the class, it makes no fucking sense that you'd fail the class if everyone else got a 91%

absolutely nonsensical, a 90% is an A, that's the only consistent way to do it

[Anonymous No.9078589]

So if you got a 1%, you got bumped up to 31% and if you got 0% you stayed at 0%? That seems kind of bullshit.

[Anonymous No.9078595]

>>9078589
meant to reply to >>9077862

[Anonymous No.9078600]

>>9077825
teachers do this because they're insecure with the difficulty of their class, there's no other reason. premeditating failing 20% of the class before anything even happens is absolutely brain-dead

[Anonymous No.9078605]

>>9078589
Yep.
Tbh if you didn't get 1% you probably deserve to get zero.

[Anonymous No.9078929]

>>9078523
>Pretend to be a genius
>Concerned with marks rather than mastery of material
Bad news for you

[Anonymous No.9078949]

C- means failed class?

[Anonymous No.9078959]

>>9078949
no. C- is slightly below the average.

he means that the entire grading is centered around the students themselves, so only like 2% of people get an A, and only like 20% get a B or higher and 50% get a C or higher, 75/85% get higher than a D, and like 15/25% fail

this is all determined before the class starts. some teachers do it like this. so your entire grade is centered around how other people do.

a C could theoretically be a 95% if that was the class average. unlikely, but this is why it's nonsensical to do it like this.

some people will call you a brainlet for hating this system

[Anonymous No.9078962]

>>9078600
This

[Anonymous No.9078964]

>>9077825
the engineering dept at my uni uses a bell curve w/ average at C+, fucking gay

I got a 95% raw score in a class and got a D+

thank god I'm in physics which doesn't use retarded curving

[Anonymous No.9078971]

>>9077825
Yes, I fucking love socialist grading schemes

[Anonymous No.9078977]

>>9077893
>>You had an hour long session every week with the professor where you orally presented your solutions and talked about any problems you had in the class one-on-one
Your professor spent ONE HOUR PER STUDENT PER WEEK on this one class?

[Anonymous No.9078998]

>>9078512
More like
>wow, engineers actually work with challenging material instead of marhfags' pretty sudoku puzzles

[Anonymous No.9079009]

>>9078964
This has to be bullshit. There's no way the average of the class was well above 95%.

[Anonymous No.9079032]

>>9078959
sounds retarded
when you do a class, the grade you get is a way of telling if you are able to pass the class or not. Coparing your students to the rest of the class is pointless

[Anonymous No.9079145]

>>9078977
He said it was a topics course, which means like 5 people max.

[Anonymous No.9079947]

>>9078035
>he's unable to be in the top 2%

[Anonymous No.9079954]

>>9078436
>he thinks the HYMPS don't have massive grade inflation

[Anonymous No.9079961]

>>9077825
Not really, especially if the question count is low.

The only "curve" I saw was a uniform shift of all grades so that the highest scorer on a given test was shifted to 100%

[Anonymous No.9080067]

>>9078998
Even more like
>wow half the class knew it was all a bullshit waste of time and they're gonna pass anyway so they got wasted in their dorms the night before, as usual

[Imnot Arobot No.9080089]

Grading "on the curve" assumes that class performance will be
distributed as a bell-curve. Is there any basis in fact for that assumption?

[Anonymous No.9080222]

>TFW you get a 90%+ on a hard test, fuck up the curve for everyone else, and they hate you for it

[Anonymous No.9080230]

>>9080089
It's pretty common, but you're right, it's not guaranteed. It's also pretty typical to see bimodal distributions or long-tailed distributions.

[Anonymous No.9080235]

>>9077869

> ywn be judged before an all knowing based professor

[Anonymous No.9080384]

>>9080235
>ywn bang your professor for an A

[Anonymous No.9080454]

>>9080235
not in this life

[Anonymous No.9080655]

>>9077825

Bell curve grading doesn't exist here. Still graduated with a 9,6 of 10 average.

[Anonymous No.9080691]

>there's no such thing as value
No, I do not like Bellcurve grading. Being the top 1% of a group of people unable to do simple math does not mean you get to ace math.

[Anonymous No.9080727]

No, because it fails to measure individual competence, which is the purpose of grades. If you get stuck in a class of geniuses, you will get a poor grade even if you have sufficient mastery, and vice versa for a class full of retards. Both are completely out of your control.