Curving a Grade - A simple and Fair Formula

Another quick www.OptimizeYourTeaching.com Idea. Curve assignment scores super fast while saving the smart kid from getting jumped on the playground. Check out this fair and simple grade curving formula.

[@WHATWEKNOWproduction]

David Spade just explained how to curve grades!

[@therichest6103]

Curving does not always cause grade inflation. It depends on the teacher. I have taken incredibly hard college classes where I and other students study dilligently and the class average is lower than 50 percent. A test can be as easy or hard as a teacher wants to make it. Some teschers give really hard tests covering many inticacies of a subject then curve the grades while others just give easier tests only covering the basics.

[@rahulgarg9328]

This is brilliant actually. Worked well for me. Thank you!

[@OptimizeYourTeaching]

Thanks for the feedback! Levels 2, 3, and 4 to come as the year unfolds and the lessons are needed. Next year, they should all "be in the can" and ready to go. FYI, send me notice of any mistakes I make, and I'll make youtube annotations to correct it and give your class props. Good luck on the rest of your year and please feel free to send me suggestions for other curriculum you'd like to see developed into ready-to-eat, flipped lessons! Sincerely, Andy Harrison

[@MatuszDawn]

Thanks for this!! An exam I created from test banks has gone off the rails (questions I deleted from the bank are still given to students somehow???) and I may need to curve. Definitely going to use this method!

[@alexs6250]

Thanks Andy. That was helpful

[@farouqalsalih619]

So basically, people who score high get rewarded the most, while people who score low get rewarded the least. In this method, William’s score improves by an additional 8.5%, while John’s improves by an additional 24%. Not saying that we should go out of our way to reward lower scores, but this curve works by pulling up the higher averages more to reach the desired 80% than the lower scores. It should either be the other way around, or have a constant added to everyone’s grades that pushes it to the 80%. The latter way won’t favor lower scores or higher scores.

[@keith2002a]

Here's hoping this gets to you on a nearly 5 year old video, but I'm interested in what software you are using to create this video.

[@OptimizeYourTeaching]

Thanks for the feedback.

[@jonathansefcik473]

The biggest curve I've ever seen in my life was for my friend's chemical engineering thermodynamics course (I didn't take it; I just heard from him). Apparently the class average was a 26, and the registrar curved the class so that a 26 became an 80 and a 5 became a 60!!!! I think its ridiculous but supposedly that's what happened.

[@2ndintelligentWorld]

that's just a terrible exam then. exams shouldnt be so difficult that it doesn't accurately reflect what it's supposed to measure

[@MrSkyboy997]

secondintelligentWorld So many STEM courses in my school are like this.

[@ivault3]

sounds about right. LOL

[@ivault3]

this is how my exams were

[@carltoncoleman454]

LOL! I kind of like this curve though as hilarious as it is. It just goes to show you that a number by itself means nothing and can be translated to mean anything. It kind of reminds you of how standardized tests are. It sounds like the registrar likely used a linear curve. A 47 in this case is likely a 100.

[@Dann0343]

I realize this is and old video, but percent increase of John (48) vs William (17), is kind of unfair in the second method. If a class average of 80% is desired, it would be more beneficial to the students that scored lower to curve to the second highest grade instead (+8). This would instead make John's grade 48+8 /50 (112%) instead of 48/40 (120%). William will now receive a 17+8 /50 (50%), instead of a 17/40 (42.5%). A quick calculation determines that increasing everyone's grade up to the second highest mark indeed brings the average up to 80%. The method described here is effective and simple but it seems like it will "flatten" the bell curve, rewarding the highest scoring marks more than the rest. Additionally, if marks above 100% are an issue, you can consider the extra marks as work half. Ie Johns 48+8 will become 48+2 to bring him up to 50, plus an additional 6*50%, thus making his total grade 53/50 (106%). It doesn't entirely mitigate the greater than 100% problem but it makes the possibility of a grade exceeding the maximum less likely. tl;dr curve to Sally's grade (+8), it will benefit William more and John less

[@carltoncoleman454]

Yeah it's more of an incentive based curve than anything. It could be used to make students who perform well have a cushion when they don't do so well on other assessments. Also a low score such as a 0% would not impact the grade nearly as much since the possible points are lower.

[@Pete-Prolly]

"Hi, my name is Pete & I'm a Math-head, I started doing Math in grade school, by Jr. High I was doing Math everyday; lots of it. I didn't have time for anything else. Then, in college I moved on to the hard stuff...(Calculus.) I guess I was looking for the limits, noone ever told me Trig was a gateway function; we were all doing it..."

[@LumenOP]

nice video! thanks!

[@bencole2884]

I've been using this curving method since I found it a couple years back, but I just noticed something undesirable: it brings the highest grades up the most and the lowest grades up the highest. To achieve a similar affect, you can simply add the difference between the actual and desired class average to each grade.

[@ratspice8959]

Thanks David Spade

[@brockmccain5270]

To your point - if the class percentage is 64% -- there is a problem with the students understanding of the information (regardless of fault). I'm not sure the answer to this should be lowering the bar for all students involved -- but it's 1 possible solution I suppose.

[@dhiraj027in]

Come from a different schooling/grading system. So please explain why would you do this, just to bring the class up to 80%?

[@teresatalerico7596]

I love this.

[@Anything13265]

Does a curve mean that all students get the same grade

[@amberhudson2033]

Very helpful. I will be using this method.

[@basileducerf1827]

Hi Mr. Harrison, Can you explain why you'd like to change your class average please? I am not sure to understand your point on changing ratings or anything about results. Thank you to reply for us, teachers in Europe, who just don't transform student results or exam ratings. Sincerly yours.

[@pigmier]

I wish you would have added an easy way to plug this into an excel sheet for us non-math-heads. Kinda hard to apply this the way you end it.

[@acmbendorf]

Really can't call this fair. While most curving rewards the lowest and that is distrubing, this rewards John (highest score) with an increase in grade by 21%, while William (lowest score) only sees an 8.5% increase in his score relative to the total. This actually "punishes" (comparatively) the lowest student when you compare the change in the point adjustment based on percentage. If this was an increase in wages the rich would be getting much richer.

[@carltoncoleman454]

Yes this is a merit based curve. Had the curve adjusted the points earned instead of points possible, everyone would get the same curve but the highest score would be a curve breaker unless scores above 100% are allowed.

[@vkak1]

Which is unfair why? If William was not prepared to do good, and screwed his exam up, then why should he get more free points? John did exceedingly well, and should retain that merit. Curve grading is bullshit to begin with. It’s academic socialism.

[@bobman929]

@@vkak1 you need curve grading to account for different lectures, uni's and unforseen circumstances. If graduates of one year scored better than the next because of these then it's unfair. It matters if they go got the same job etc.

[@danibaba7058]

isn't that fair?if you work more you get rewarded more.Your logic is flawed.

[@c0r5e]

@@vkak1 john has A regardless. Struggling students often do more to get less and no one studies less because they will get a curve

[@OptimizeYourTeaching]

Good, but what happens when you have a student with stellar performance. Adding 10% to an aced assignment give a heck of an inflated grade AND you have to go through EVERY grade to make corrections, rather than just make a single adjustment to the total points possible.

[@jacobhubbard9266]

I understand this is a 10 year old video but videos like this confirm to me why grades in general are bad way to teach students. I understand that we have to have grades as part of the job, but why make this more difficult than it needs to be. I see no benefit to grading on a curve like this other than to force students to compete for a grade rather than getting them to focus on their learning. This is ridiculous.

[@kendrickbrenton2164]

Not unfairness but idiots passing when they should not грустный

[@kyriakospetakos3154]

in that case why do not we add 8 points to everyone as long as we do not exceed 50?

[@brockmccain5270]

Grades are imaginary scores now. Lol

[@d33san]

based from the results, what passing percentage I am going to use?

[@williamgosset1]

What was the calculation to get the 49/40? Didn't explain that

[@TobyBW]

This is so stupid. It give you the grades you want instead of what's actually happening.

[@seanm2818]

I take it you aren't a statistician of any sort. The numerical grade is meaningless if other factors are unaccounted for. A 70% is an outstanding grade on an AP Calculus, a solid score of 5. A curve can get a more meaningful representation of the students' ability.

[@richardkoeln]

I take it you aren't a teacher of any sort. If someone only scores 70% on a test you don't represent their performance better by fudging the numbers until it looks like they have 100%. And if 70% is the best students can do on a given exam, you either suck at teaching or at test conceptualization.

[@carltoncoleman454]

Numbers by themselves don't mean anything. Every score when it comes to evaluations or any other marking standards is a curve of some sort. None of the grading systems are uniform across the world so this is not stupid. A 70% everywhere is not a C! Some countries make exams difficult so that the raw scores are low even though that may sound harsh but then they curve so scores will meet a certain distribution. A 25% in some cases can be considered a B or even an A while a 90% could be a C in others.

[@wilterfilho]

I’m watching this because I’m from Brazil but I’m an exchange student here in the US and apparently I just set the curve for my International Marketing exam and got the full mark because of that... but in Brazil we don’t use this, so I’m trying to understand why I did so well 😂

[@NirousPlayers]

Mano to tentando entender pq porras eles fazem isso com as notas Wtf.

[@69adrummer]

And we get dumber as a society. Well done John.

[@alanimations]

Why would you ever do this

[@tracibrown-perkins8088]

Does this formula work the same if the test was worth 100 point?

[@agn003]

Am I missing something? How can John receive a grade of 120%? He was previously on 96%, so how do you tell a student they achieved 20% more than what is mathematically possible?

[@Ncloud]

It means he overachieved considering the difficulty of the exam. That's what the extra 20% means.

[@danibaba7058]

@@Ncloud nah bruh he got 96% not 120% the extra was just gifted to him.

[@joep7226]

Enemy of the state...lol

[@ba_brandon]

John still made the class average high so people probably would be mad at him still. In the class average, you shouldn’t include the highest score and lowest score since they might drag the average too much

[@danibaba7058]

or leave the goddam marks as they were, not change it cuz u look bad as a teacher

[@Unv12sL]

I hope my data structures professor implements this formula ;D

[@jbelle7133]

If you earned the F you get an F, just like an A!

[@damonhart8595]

I don't understand how John got punished in the first method ! he got 48/48 that means 100% , how is that a punishment ?

[@gweng123]

They get mad for having to do the extra work

[@damonhart8595]

Gwen Goldenberg what extra work ?

[@jnzed94]

He gets punished because the other students will get mad at him or call him out or put him on blast because the other students could have gotten more extra points if "John" had gotten a lower score.

[@gweng123]

+Damon Hart extra work to figure out the other grades

[@vkak1]

Jimmy N And that’s what’s wrong with society to begin with, asshats that cannot realize that you have to work to be successful, and rather want free fuckin shit

[@joshuawertheimereastcta896]

Except this isn't really fair. When you subtract from the total points possible, you raise the c student's score about 20%, and you only raise the bottom student's score 8%. Check the numbers...

[@nathanrice2481]

You do realize this isn't "curving", right? Curving means you force the grades to fit on a normal distribution curve. If the scores were low, then maybe spend more time looking at the assessments before the test (homeworks, classwork, quizzes, etc.). If the other assessments were not great, then take more time on the material before the test. Or I guess you could arbitrarily set the desired average so it looks like your students are great at the material. You aren't optimizing teaching, you're optimizing how your kids look and teaching them incorrect mathematics along the way.

[@HereNowOOO]

Why is giving John 2 more grades and making his mark a full one interpreted as "punishing" him?!

[@carltoncoleman454]

He is referring to the fact that many teachers will only curve to till someone gets 100%. Giving everyone two points would make him the curve breaker. Instead of allowing him to be the curve breaker, this method makes the average of the class 80% while awarding those who did well with a higher percentage.

[@danibaba7058]

@@carltoncoleman454 so what's wrong with giving everyone just 2 points???Since when is equality bad?Not to say that adding 2 points out of your ass just shows how "good" of a teacher you are.

[@carltoncoleman454]

@@danibaba7058 The most common method would be to give everyone two points. A better option would be to do rewrites for partial credit or teach it effectively enough the first time and provide study guides so you don't have to curve in the long run.

[@danibaba7058]

@@carltoncoleman454 yes but doing it the right way is harder...unlucky

[@andrewlabna]

so s l o w

[@apolonatomus]

Curving grades is a shame and an important underlying cause behind the decline and erosion of the education. Not wanting more from students equals making them happy from little, supports their unjustified self-satisfaction and self-appreciation and hurts them in a long run.

[@danibaba7058]

preach brother

[@apolonatomus]

@@danibaba7058 Are you also benefiting from curving grades?

[@danibaba7058]

apolonatomus where did you come with that conclusion?I literally said i agree with you and you’re saying i benefit from them?If i get a grade that’s my grade from my efforts not from some curve(not in uni yet and i hope we wont have this shitty ass curve)

[@apolonatomus]

@@danibaba7058 sorry, I missed your point. Keep up doing good work 👍

[@danibaba7058]

apolonatomus thank you man and np,you too!

[@mariamalnaqbi3544]

okay , what about the others ???????????????????????

[@mikeymo4]

Problem with that of course is that the students will catch on quickly and all just not study for tests because they know that if they all do crappy then they will all be rewarded and get a good grade anyways.

[@hwwwarrior90]

"If we wind up with a class average of 64 either I've done something wrong or the students have"... my classes are all scaled to 60% :0

[@whittakerdanielj]

Teachers will never take responsibility for how they teach or grade, they just sly around it with this. This is cheating the students out of their grades.

[@alcastilloa]

Shameful promotion of mediocrity

[@joshl90]

You’ve never been in STEM courses then, have you?

[@seanm2818]

Josh, generally speaking the commenters here probably haven't taken even as high as Algebra 2. Curving a grade is perfectly normal, it isn't grade inflation, in fact every standardized test gives a "scaled score," not a raw score.

[@13donstalos]

It just seems like such a sham to me, but that's probably because I bust curves for a living

[@ando1135]

so the guy who got 38 originally had a C+ but by doing what you did, his shitty performance now gives him an A...something is not right here, OH YEA! your grading system is wrong. it's basically "how can i make my class look good when theres really only 1 good student here? i know, i'll grade based on my desired percentage outcome!" this doesnt benefit students, it just makes them think they are doing better than they are.

[@mwidunn]

You didn't show how to solve for X.

[@barbietripping]

4:25

[@mwidunn]

@@barbietripping Sorry -- no. He just gives the answer. He doesn't go through the process of showing how he arrived at "x=40".

[@conquesotador]

@@mwidunn it’s a proportion, you multiply the bottom opposite so 32*100 then divide it by the number on top 3200/80 = 40 (Cross multiply then divide to solve for X) I’m not sure I explained it well, you can search videos on the subject if you need more help.

[@mwidunn]

@@conquesotador He does not explain how he got from A to B. So, . . . fail.

[@jeremykirkpatrick472]

You sound like Joe Dirt. 😂😂

[@idaho_girl]

Please people do not curve grades. Doing so ignores if a class is high-performing or low-performing. It gets the causality arrow backwards; it forces the data to be fit to an expected outcome rather than describing the outcome. And I am also disturbed by the idea of what is "fair". What we should desire to do is be accurate in our grading so that a grade is an accurate reflection of an individual student's master of the material. And as for qualification: I am an experimental physicist and professor for almost 20 years.

[@envyling2461]

Fair? I wouldn't say the Curve System is "Fair"! The Curve System would only works when the class average is sooo low like below a 50%. But what if you have a class where the majority are highly devoted, highly dedicated, & hardworking students? That unfortunate scenario, we would see it would ruin the curve. Especially, if the class average in the devoted group are of course around 90% where the lowest in the class is 85%. That means in terms of the curve, the guy who got 85% will basically Fail the course, meanwhile those who got exactly 90% are getting a C, & those who want to get an A in the course have to really push hard of nearby 98-100% to get an A. Sorry, but I rather stick with the Absolute System, where 90-100 is A, 80-89 is B, & vice versa... Absolute System works only for the highly dedicated groups. Whereas the Curve System only works for the slackers.

[@mrdorscher974]

This has made me lose more faith in the education system, ffs just teach the kids.

[@Msvip11]

i wish teachers use the new way, they all are using the old way which is unfair!

[@nothingisreal_4056]

fart noise

[@Josh54152]

The is no such thing as fairness when it comes to curving grades! Curing is the most unprofessional and stupidest thing a teacher could ever do!

[@TimTeatro]

Josh P That simply isn't true. It's just a tool that needs to be used properly, and most teachers don't. The point of statistically manipulating grades is to eliminate the effects of factors such as having multiple lecturers for the course, having multiple versions of an exam to stifle cheating, etc.

[@michaelsmoak8804]

You are literally David spade

[@pytulyn1]

ridiculous

[@williamgosset1]

so Sally and John would have greater than 40 points or an A for the test. In effect adds 16 percentage points or 8 test points onto everyone's grade. Nothing original in that. An easier way would be to obtain the difference between the desired class average and the actual class average, add that to each score and voila - you have your grade. I have a proportional system which elevates everyone's grade in proportion to the deficit from perfect. I obtain that proportion by dividing the desired deficit by the actual deficit and then "correcting" that grade student by student. An Excel spreadsheet makes that all easily possible. This obviates the need for making false assumptions about a bell curve which never exists. If the desired deficit is MORE than the actual deficit then I just obtain the difference and subtract that from everyone's scores ("curving down"). Unfortunately, if you have too many over achievers which decreases the actual deficit then this forces a downgrade to everyone"s grade. I just take care of that on a case-by-case basis. The trick to accurate testing is the designing of good questions and requiring well thought out term papers. This comes with years of experience. Again, motivated achieving students do not score by the bell curve. It is either a uniform distribution or a reverse chi-square distribution. One cannot force a bell curve on these students. Hence, most times if the class average is high, then I don't deduct anything and give all the students scored the equivalent of an A an A.

[@MSCGERECO]

Pathetic technique.

[@pranjal86able]

that is so stupid. you dont even need the formula. you are not solving the problem. how do you reward john now? and this was not mentioned to the students in the beginning so that is not fair. if you want to do this properly, map all the numbers from 0 - 50 to higher numbers such that 32 is mapped to 40. another alternative (less fair, simpler) would be to score from -20 to 50 not 0 - 50. that way you can have your higher average.

[@alexyu1236]

How the fuck is John rewarded from a 48/50 to a 80% . Grading on a curve is pure manipulation of grades if John got 48/50 then leave it as is. And teachers allowed to lower his grade ? I'd be pissed as fuck.

[@basrrf]

John Smith The class average is 80%. That doesn't mean he individually gets an 80%. He gets rewarded with a grade of 48/40, which is 120%

[@alexyu1236]

AH i see but still don't like grading on a curve to me feels like teacher some how manipulating scores by having a certain percentage of people can only get A B C and so on. To me teacher should just allow students to get what they get, if half the class gets A, then half the class gets A, a teacher job is to mark the tests that comes and not change anything to their scores. In this demonstration using the formula i feel again its like manipulating the score to what the teacher wants the average to be...

[@basrrf]

John Smith Yes, grade curves are technically manipulation of grades, but what if a teacher didn't teach a class of students as well as they had previous years before? I believe it's fair for teachers to reasonably curve their classes in order to "make up" for any "mistakes" made by them, and to be fair to the students. By the way, I'm talking about college grade curves, not high school or whatever.

[@alexyu1236]

basrrf if a teacher didn't teach as well as they did previous year then that's on then that's their mistake they will have to live with but then because they can manipulate grades so it's fair on the teachers? To me that's like trying to get away with poor teaching in a way. And I believe in no way shape or form manipulation of grades are fair for students.

[@basrrf]

John Smith Well, also take into account that many universities have multiple professors for the same class. It wouldn't be fair for students to get a professor that sucks ass compared to students that get a great professor, all for the same class.