Is it Normal to Have a 5-Point Curve Applied to Total Points on an Exam?

  • Thread starter Thread starter cue928
  • Start date Start date
  • Tags Tags
    Curves
Join the discussion
Registration is free. Start your own thread to ask a follow-up.
Sign up Log in
9 replies · 3K views
cue928
Messages
129
Reaction score
0
I had a first recently. Took an exam for which there was a 5 point curve. What surprised me was that not only was the five points added to everyone's score but also to the total points available. :confused: Is that normal? For me, it increased my grade by two tenth's of one percent.
 
Physics news on Phys.org

Discussion

cue928 said:
I had a first recently. Took an exam for which there was a 5 point curve. What surprised me was that not only was the five points added to everyone's score but also to the total points available. :confused: Is that normal? For me, it increased my grade by two tenth's of one percent.

Very strange and pretty much pointless. Does the prof have a clue about how math works?
 
I dunno. Adding the points to the bottom line kept me from moving up a grade level. I'm personally livid about it. I mean, what the h*** is the point of that? I just don't freaking get it.
 
cue928 said:
I had a first recently. Took an exam for which there was a 5 point curve. What surprised me was that not only was the five points added to everyone's score but also to the total points available. :confused: Is that normal? For me, it increased my grade by two tenth's of one percent.

It's an odd way to curve, which basically amplifies the lowest scores the most... an individual legitimately getting a 0% on the original exam (i.e. attending but getting no points) would get a 4.8% score (5/105), while a student getting a 100 on the original exam would still only get a 100% (105/105). It seems a little odd, but I would think that the instructor knows what he/she has chosen to do and has some reason for it.

I've done the opposite before, amplifying the highest score most by scaling an exam's total points to the highest score. Say the highest score was a 90%. That student would then get a 100% (90/90), but a student legitimately receiving a 0% would still receive a 0%. I actually only did this my year of teaching, when I was having a bit of trouble scaling the length of the exams. I figured the top students just had a problem with time, while the students were just pretty much clueless, even though they had tons of opportunities to do at least something right.

Now that I'm more experienced, I think I'm better at writing a reasonable exam (with reasonable expectations), so I only curve on the occasion of a typo that might have confused students (and then I'll curve in a standard way, by adding a certain number of points to all the students scores). If I think the scores during the course of the term aren't good (but my expectations are reasonable), I'll tend to include a bonus on the final to give students a chance to redeem themselves.
 
Such a curve will raise low grades more than high grades. That might well have been the intent.

I also think that your anger is misplaced. Surely if you got more points you would have gotten a higher grade.
 
The test was written as a 95 point test. Beyond a certain point, it still only makes a fraction of a percent improvement. Say someone got a 50/95, which is 52.63%. 55/100 is a (round numbers) 3% improvement. But yet if you got an 80/95 (84.2%), you only get a whopping 8 tenth's of a percent point improvement. To me, that is not right because it rewards the people who had a failing grasp of the material more so than people who worked hard and actually "got" the material.

At some point, you're just playing with numbers. In physical chemistry at my school, for example, there is an "all or nothing" style to grading. Compared to our math department, which gives partial credit, in p-chem, it's all or nothing. The advisor went on to say there is a significant curve at the end because "there would be a huge outcry if 3/4 of the class failed." So in musing about this to a friend who pointed out that the whole thing seemed rigged. The tests, he claimed, were written and graded in such a way that most people (3/4), would fail on their merits. Yet they allow what amounts to a monstrous curve to save everyone at the end. Something doesn't seem right there and I would have to agree. It's all just a numbers game. But I digress...
 
Last edited:
I don't see the point of grade curving at all. I never heard such a thing at my university and sometimes all the people who took the final exam (worth 100% of your grade for the course) fail. I've seen 5 people taking a 3rd year course and they all failed, some by 1 point/10.
 
Conversation's morphing a bit but I guess that's okay. Some people will succeed and unfortunately, some will fail. As someone who wants to teach, what troubles me the most is when underperformance results from poor expectations laid out in class. I returned to school for a different degree after a short stint in the workforce. The best bosses I had were those who clearly laid out their expectations. This of course doesn't mean the tasks assigned were easy. Many assignments pushed my level of programming skills much higher had it not been for them. Nonetheless, the instructions were clear.

So being in school now, having had those experiences, causes me to look at things somewhat differently. Perhaps not surprisingly, it is often not favorably.
 
cue928 said:
Conversation's morphing a bit but I guess that's okay. Some people will succeed and unfortunately, some will fail. As someone who wants to teach, what troubles me the most is when underperformance results from poor expectations laid out in class. I returned to school for a different degree after a short stint in the workforce. The best bosses I had were those who clearly laid out their expectations. This of course doesn't mean the tasks assigned were easy. Many assignments pushed my level of programming skills much higher had it not been for them. Nonetheless, the instructions were clear.

So being in school now, having had those experiences, causes me to look at things somewhat differently. Perhaps not surprisingly, it is often not favorably.

I agree.

To me, pretty much any grading scale is fair as long as the expectations are made clear.
 
kylem said:
To me, pretty much any grading scale is fair as long as the expectations are made clear.

That sums it up pretty well. Personally, I think profs should make a good effort to not have to curve, but it happens sometimes. If a prof is consistently curving so students with 50% are getting A+, then they are doing it wrong.

You could always leave a copy of Krantz's "How to teach Mathematics" on his doorstep :-)

https://www.amazon.com/dp/0821813986/?tag=pfamazon01-20

Joking aside, it is a good book for anyone thinking they might want to teach a mathematical subject.
 
Last edited by a moderator: