# Unfair Grading with floating bar?

Discussion in 'General Education' started by atomic, Feb 9, 2012.

1. ### atomicCompanion

Joined:
Jan 29, 2009
Messages:
183
0

Feb 9, 2012

Would anyone think it unfair to use the average(mean) scores for a test to set a passing score?

My father talked of professors using a bell curve to grade. There was always a certain number of failures even if the lowest grades were in a typical C range. He said they were always fighting to "not be in the bottom of the class."

3. ### mrachelle87Fanatic

Joined:
Aug 30, 2006
Messages:
2,815
53

Feb 9, 2012

My son is in a college class where the professor uses a bell curve. The second week of class, he gave a practice test. My son only missed one out of 75. The professor pointed to him and said, "Here is the guy you all need to hate. He is going to mess up your bell curve." My son thought it was funny at first, but then when others started making rude remarks because he always has his homework submitted on time and always 100%, it started bothering him. He took a big test yesterday (worth 1/3 of the grade) and I asked how he thought he did. He told me that he missed at least one, but that he had thought about blowing it so that he wasn't responsible for the curve. He said that he decided that he was going to keep his A and not worry about it. I am frustrated that any educator would basically draw a target on my son's back. And I hope he sets the curve high!

4. ### TeachOnHabitué

Joined:
Jan 28, 2012
Messages:
804
0

Feb 9, 2012

The schoolhouse would look like a reenactment of the original Bastille Day, only it would be a crowd of parents.

(and it is a bad idea)

5. ### a2zVirtuoso

Joined:
Sep 16, 2010
Messages:
6,347
2,231

Feb 9, 2012

When the mean or a bell curve is used to determine the passing grade it only serves the teacher. The grades are supposed to represent mastery of the material, not how students performed against one another.

If students consistenly score with a mean of 65% it shows either the test is in error for testing mastery or students aren't mastering the subject matter.

By using this method it is possible to have a class mean of 25% and these students get a C. We could get to the point where kids just show up and get Cs or As for that matter if all end up with zeros.

6. ### atomicCompanion

Joined:
Jan 29, 2009
Messages:
183
0

Feb 9, 2012

So why is it ok for states to do this for high school exit exams?

7. ### TeachOnHabitué

Joined:
Jan 28, 2012
Messages:
804
0

Feb 9, 2012

States don't have ambitious parents who expect Johnny to sail into Harvard Yard on an ocean of A's.

8. ### atomicCompanion

Joined:
Jan 29, 2009
Messages:
183
0

Feb 9, 2012

That is true, but a graduation exam that the scoring is done in a way that not everyone can pass seems a bit unfair to me.

I'm not 100% sure I've got it right that it's how the grading is calculated, but I think it is.

I'm trying to figure out how every year, no matter how much more we do and no matter how much we teach, it's always around the same number of students that don't make it.

I'd love to see this statistic for the surrounding schools for the last ten+ years.

9. ### a2zVirtuoso

Joined:
Sep 16, 2010
Messages:
6,347
2,231

Feb 9, 2012

I don't know if all states do this. Do you have a link to show this is done in your state?

Most states that I know of use criterion-referenced tests that do have cut off scores that are not based on the bell-curve. They are based on what the state has determined is the right amount of knowledge to say the student is 'proficient enough' to graduate. It is possible that 100% of students will pass this criterion-referenced test, but not likely.

There are national tests such as the Iowa tests that is norm-referenced and is used to compare a student to the nation. With a norm-referenced test there is no passing or failing, just where a student falls within the bell curve from norming cycle to norming cycle. In theory, a norm-referenced test will not distinguish between an illiterate society and a well educated society based on the results. The curve can move either way, up or down, with each normalization, which in some cases is every 5, 8, or 10 years or so. The bell can rise or flatten based on the spread between the top and bottom performers. It says little about what they know but says something about how they compare to others.

State tests are typically criterion-referenced. There is a set of standards and students must know a set amount of information.

10. ### catnfiddleModerator

Joined:
May 8, 2008
Messages:
8,584
1,700

Feb 9, 2012

I had a high school teacher who, in place of a bell curve, had the HAS (Highest Average Student). The score of the HAS would become the 100% for the class. It was always my friend Josh and we all knew it. If the HAS scored 90 out of 100 points on a test, everyone's score was calculated as if it were a 90 point test instead of 100. Nobody had a problem with this whatsoever. It wasn't a curve, but an ease of how many points each assignment was worth. If anything, everyone wanted to be Josh's lab partner.

11. ### AliceaccMultitudinous

Joined:
Apr 12, 2006
Messages:
27,534
6

Feb 9, 2012

It depends on the purpose of the testing.

On tomorrow's geometry test, I'm trying to see who has mastered the material. Ideally, everyone is, at this very minute, busy studying the 12 Area formulas on the test, and everyone would get 100%, signaling that they had all mastered the material. Lowering someone's grade would serve no purpose-- the intent behind the test is to guage who has mastered the material.

A normal distribution is used to illustrate something else entirerly. Things like height and intelligence follow a normal distribtion. In that case, 68% of the sample falls within one standard deviation of the mean. Let's call them the "average" group. The further you get from the mean, the fewer people you have.

Joined:
Jun 1, 2008
Messages:
2,019