Baltimore to ban grades lower than 50%?

[Emily's Cat]

A failing score does not represent 60% of the scores achieved so all of your "mathematics" is irrelevant.

Even a poor student (or a student who is having a bad day) is likely to score considerably more than 0% on an assessment task so a failing test result will still add to the aggregate. Only students who make no effort whatsoever risk getting a zero there is no good reason why such a slacker should be rewarded for that.

Look - if a good student has a bad day and gets a C (75%), but previously had an A (95%), their grades average to a B (85%). It's nicely and evenly in the middle.
A student who's having a rough time who gets a low F (say 20%) but previously had a middle C (75%) will have an average of an F (47%).
A student who starts with a B average (85%) and gets a low F (20%) ends up with an F (52%).
A student who starts with an A average (95%) and gets a low F (20%) ends up with an F (57%)

A low F brings an aggregate grade down by far, far more than any other grade can affect it. This isn't really up for debate - this is how the math works.

Do you think it's reasonable that a badly flubbed test can bring an otherwise good student's grade all the way down to an F? Remember, this isn't solely about missed assignments.

Let's try different tack. Scrap percentages.

• Would you support a grade system in which A is worth 4 points, B is worth 3 points, C is worth 2 points, D is worth 1 point, and F is worth 0 points?
• Would you support a grade system in which A is worth 9 points, B is worth 8 points, C is worth 7 points, D is worth 6 points, and F is worth 0 points?

 

[Emily's Cat]

So what is the point of gifting a total slacker 50% if their "grade point" remains 0.0 regardless?

This is where we have a disconnect. It's not about giving the slacker 50%. It's about giving a diligent but struggling student the mathematical opportunity to pass the class, instead of consigning them to a score that is impossible to improve.

 

[fuelair]

And, since Baltimore started this feast, here is theirs:
www.bcps.org/pdf/grad_requirements.pdf

 

[fuelair]

OK in Dogpile.com this took me straight to the list but hitting it from here it is not ( it says missing. but when I went to it after that on the Dogpile list it went to it just fine). No idea what the problem is, but I searched and found it using Baltimore Public High School Graduation Requirements and the site shown on my previous post took me to it just fine so I copied it to here but it says 404 error from here. Computer person may know why, I am not one.

 

[fuelair]

Look - if a good student has a bad day and gets a C (75%), but previously had an A (95%), their grades average to a B (85%). It's nicely and evenly in the middle.
A student who's having a rough time who gets a low F (say 20%) but previously had a middle C (75%) will have an average of an F (47%).
A student who starts with a B average (85%) and gets a low F (20%) ends up with an F (52%).
A student who starts with an A average (95%) and gets a low F (20%) ends up with an F (57%)

A low F brings an aggregate grade down by far, far more than any other grade can affect it. This isn't really up for debate - this is how the math works.

Do you think it's reasonable that a badly flubbed test can bring an otherwise good student's grade all the way down to an F? Remember, this isn't solely about missed assignments.

Let's try different tack. Scrap percentages.

• Would you support a grade system in which A is worth 4 points, B is worth 3 points, C is worth 2 points, D is worth 1 point, and F is worth 0 points?
• Would you support a grade system in which A is worth 9 points, B is worth 8 points, C is worth 7 points, D is worth 6 points, and F is worth 0 points?

Prefer and would choose first option.

 

[Emily's Cat]

Prefer and would choose first option.

So would I. I'm quite interested in which psion would support. Perhaps he is indifferent, I didn't make it an either or. I genuinely want his preference on it, and hopefully some reasoning as to why.

 

[mijopaalmc]

I don't want to speak for the whole world here but where I work that's an F.

In my experience as a student in the Pacific Northwest, F has always been 60-67%.

 

[psionl0]

A student who's having a rough time who gets a low F (say 20%) but previously had a middle C (75%) will have an average of an F (47%).
A student who starts with a B average (85%) and gets a low F (20%) ends up with an F (52%).
A student who starts with an A average (95%) and gets a low F (20%) ends up with an F (57%)

A low F brings an aggregate grade down by far, far more than any other grade can affect it. This isn't really up for debate - this is how the math works.

Wow. That was quite educational.

Every test where a student gets an F is worth 50% of the total marks for the course no matter how many tests the student does.

Silly me. I would have thought that that getting an F wouldn't change the percentage of the total marks that the test contributed to. If a student got 20% in a test that was worth 20% of the total marks then in my ignorance, I would have thought that your scenarios would work out as follows:
The middle C student (75%) would end up with 64% for the course.
The B average student (85%) would end up with 72% for the course.
The A average student (95%) would end up with 80% for the course.

Sure, that poor performance adversely affected their final grades (why shouldn't it?) but nobody ended up failing the course.

Let's try different tack. Scrap percentages.

• Would you support a grade system in which A is worth 4 points, B is worth 3 points, C is worth 2 points, D is worth 1 point, and F is worth 0 points?
• Would you support a grade system in which A is worth 9 points, B is worth 8 points, C is worth 7 points, D is worth 6 points, and F is worth 0 points?

This is a different issue altogether. What you are describing is "ordinal" data - data that can be ranked but not scaled. The "points" add no further information about the student than you got from their grades.

It is quite amusing really. As math teachers, we teach students that you can not add, subtract or "average out" ordinal data but the bureaucrats expect us to do exactly that with student grades.

 

[fuelair]

Wow. That was quite educational.

Every test where a student gets an F is worth 50% of the total marks for the course no matter how many tests the student does.

Silly me. I would have thought that that getting an F wouldn't change the percentage of the total marks that the test contributed to. If a student got 20% in a test that was worth 20% of the total marks then in my ignorance, I would have thought that your scenarios would work out as follows:
The middle C student (75%) would end up with 64% for the course.
The B average student (85%) would end up with 72% for the course.
The A average student (95%) would end up with 80% for the course.

Sure, that poor performance adversely affected their final grades (why shouldn't it?) but nobody ended up failing the course.


This is a different issue altogether. What you are describing is "ordinal" data - data that can be ranked but not scaled. The "points" add no further information about the student than you got from their grades.

It is quite amusing really. As math teachers, we teach students that you can not add, subtract or "average out" ordinal data but the bureaucrats expect us to do exactly that with student grades.

That is mostly why bureaucrats suck, they are hired to do that sort of ****.

 

[psionl0]

I'm sorry, I must be missing something here. Teachers in many places aren't free to just throw up their hands and call a kid a slacker. Presumed responsibility ultimately will fall on the teacher. If I can't get a kid to pick up a pencil that is on me, not them. Once they turn 18, all bets are off but in the meantime students are presumed not to be responsible for their own lack of motivation. Therefore teachers are interested in trying harder to motivate their students.

Yes, there are some bad teachers out there but not every bad student is the fault of a teacher (or do you subscribe to the "Boys Town" theory?). More harm is done to the average student by the policy of allowing bad students to continue their mission to cause as much disruption in every class room as possible than anything else.

If you really think 50 percent is a passing grade, that explains quite a bit.

Now we are getting into repetition territory.

The difficulty of a test is (or should be) set such that a satisfactory student will easily pass it. There is nothing magical about the number that the pass mark is set at. The test difficulty is simply set according to the pass mark level.

I have had university lecturers who's policy was to make a test as difficult as possible so that they could sort out the brainy students. The pass mark on those tests necessarily had to be less than 50%.

 

[psionl0]

If a student got 20% in a test that was worth 20% of the total marks then .....
The middle C student (75%) would end up with 64% for the course.
The B average student (85%) would end up with 72% for the course.
The A average student (95%) would end up with 80% for the course.

Just to continue with this thought experiment, if a high D student who was averaging 69% thus far got 20% for the same test then his average for the course would be 59.2% which is a fail if the pass mark was set at 60%.

Such a student would rightfully feel cheated since another 5% in the test would make the difference between passing and failing the course. There is a strong case for giving that student a make-up test. He doesn't even need to pass it. It could be stipulated that as long as he got at least 30% for the test, he would get a conceded pass for the course. (And don't tell me that this is too much work for teachers. They have to jump through much more arduous loops to satisfy the bureaucracy).

OTOH Pete the party animal who hasn't passed a single test throughout the course has no chance of boosting his average enough to pass the course in a make-up test and probably doesn't deserve the chance to either.

 

[Emily's Cat]

Wow. That was quite educational.

Every test where a student gets an F is worth 50% of the total marks for the course no matter how many tests the student does.

Silly me. I would have thought that that getting an F wouldn't change the percentage of the total marks that the test contributed to. If a student got 20% in a test that was worth 20% of the total marks then in my ignorance, I would have thought that your scenarios would work out as follows:
The middle C student (75%) would end up with 64% for the course.
The B average student (85%) would end up with 72% for the course.
The A average student (95%) would end up with 80% for the course.

Sure, that poor performance adversely affected their final grades (why shouldn't it?) but nobody ended up failing the course.

You know, I kind of thought it was obvious that I was generating illustrative examples from a total of two items. I'm starting to think you'll do whatever you can to avoid addressing the actual mathematical impact that is at issue here.

This is a different issue altogether. What you are describing is "ordinal" data - data that can be ranked but not scaled. The "points" add no further information about the student than you got from their grades.

It is quite amusing really. As math teachers, we teach students that you can not add, subtract or "average out" ordinal data but the bureaucrats expect us to do exactly that with student grades.

I see that you declined to give your preference. Why?

 

[Emily's Cat]

Just to continue with this thought experiment, if a high D student who was averaging 69% thus far got 20% for the same test then his average for the course would be 59.2% which is a fail if the pass mark was set at 60%.

Such a student would rightfully feel cheated since another 5% in the test would make the difference between passing and failing the course. There is a strong case for giving that student a make-up test. He doesn't even need to pass it. It could be stipulated that as long as he got at least 30% for the test, he would get a conceded pass for the course. (And don't tell me that this is too much work for teachers. They have to jump through much more arduous loops to satisfy the bureaucracy).

OTOH Pete the party animal who hasn't passed a single test throughout the course has no chance of boosting his average enough to pass the course in a make-up test and probably doesn't deserve the chance to either.

Pete the party animal has no chance to pass the class if he got 50% on all of his failed assignments either. So what's the problem? Do you feel the need to not just fial, but to utterly bury and punish a student who doesn't live up to your expectations? Is it a case of not just failing those you deem unworthy... but making sure that there view toward education is completely destroyed and that they know that they're completely worthless in your eyes?

 

[psionl0]

You know, I kind of thought it was obvious that I was generating illustrative examples from a total of two items. I'm starting to think you'll do whatever you can to avoid addressing the actual mathematical impact that is at issue here.

You most certainly did not make it clear that you were only considering examples of courses where there were only two equally weighted assessment items. It is just too unusual.

OTOH it is not so unusual for a course to have a final exam that is worth up to 50% of the course total. Of course, if you are going to give that much weight to a final exam then you can't complain that it has a disproportionate influence on a student's final marks. The solution is to not give so much weight to just one assessment item.

I see that you declined to give your preference. Why?

Is the concept of "ordinal data" too complicated for you to grasp? If the "points" are not being added together then if makes no difference what numerical value you attach to each grade provided that A > B > C > D > F. All such scoring systems are equally valid and a fail is a fail is a fail.

 

[Roger Ramjets]

We have a curriculum that the broad range of humans are expected to be able to do.

So they should all pass the tests with high marks. But we don't want that. We must stratify them to maintain our social structure and ensure a steady supply of cheap labor.

 

[mijopaalmc]

Is the concept of "ordinal data" too complicated for you to grasp? If the "points" are not being added together then if makes no difference what numerical value you attach to each grade provided that A > B > C > D > F. All such scoring systems are equally valid and a fail is a fail is a fail.

Except it is not really ordinal data. "A" has an absolute, numerical value of 1 not relative, ranked value of 1st. This is important because, as you said, ordinal data evaluated with the regular of arithmetic operations, while quantitative data can.

One of the problems, as I see it, is that the estimator of a student's overall performance is usually a weighted mean degrades obtained in the class. The mean is a biased estimator that is heavily influenced by outliers in the data set. Therefore, basing a final grade off such a biased estimator can disastrous for a student who receives a small number of low grades relative to the number of high grades (or, equivalently if they do poorly on the highly weighted assignments).

 

[Emily's Cat]

Y
Is the concept of "ordinal data" too complicated for you to grasp? If the "points" are not being added together then if makes no difference what numerical value you attach to each grade provided that A > B > C > D > F. All such scoring systems are equally valid and a fail is a fail is a fail.

Yeah no. You decided they were ordinal. You also decided that ordinal data can never be aggregated (even though this is exactly what is done with GPA). You're sidestepping the question. Make the percent of correct questions be whatever you want it to be for each grade range. Then each grade is worth a specified number of points. Not ordinal values, points. We're not creating an artificial ranking, we're literally assigning points, you get that right?

Of the two schemas I listed, what are your thoughts on them - is either preferred above the other and why?

 

[Emily's Cat]

So they should almost all pass the tests with high marks.

Corrected for accuracy.

But we don't want that. We must stratify them to maintain our social structure and ensure a steady supply of cheap labor.

I suppose if you get your kicks foisting absurd commentaries on people, knock yourself out.

 

[quadraginta]

<snip>

Silly me. I would have thought that that getting an F wouldn't change the percentage of the total marks that the test contributed to. If a student got 20% in a test that was worth 20% of the total marks then in my ignorance, I would have thought that your scenarios would work out as follows:
The middle C student (75%) would end up with 64% for the course.
The B average student (85%) would end up with 72% for the course.
The A average student (95%) would end up with 80% for the course.

Sure, that poor performance adversely affected their final grades (why shouldn't it?) but nobody ended up failing the course.

<snip>

I don't think you and E.C are all that far apart. You seem to be talking past each other.

I hadn't gotten the idea that E.C. was necessarily advocating that someone who did absolutely no work at all should get 50% credit, but rather that when someone's test score was converted to an F that it essentially eradicated any partial credit they might have earned.

You seem to be saying that as long as they got that partial credit then the grading system would not be unfair.

These two positions are not totally incompatible.

<snip>

Is the concept of "ordinal data" too complicated for you to grasp?

<snip>

Okay. That was uncalled for.

E.C. has done nothing to deserve that sort of scorn and condescension.

You do nothing to advance your arguments by resorting to insult.

 

[mijopaalmc]

Except it is not really ordinal data. "A" has an absolute, numerical value of 1 not relative, ranked value of 1st. This is important because, as you said, ordinal data evaluated with the regular of arithmetic operations, while quantitative data can.

This should read:

Except it is not really ordinal data. "A" has an absolute, numerical value of 1 not relative, ranked value of 1st. This is important because, as you said, ordinal data cannot be evaluated with the regular of arithmetic operations, while quantitative data can.