4 ms·

I don't really like being overly strict. I think professors and especially teachers have gone too soft on students (often forced to do so by admin), but I also

I don't really like being overly strict. I think professors and especially teachers have gone too soft on students (often forced to do so by admin), but I also find being too strict means the grade is now less about the topic and more about minor nuances that aren't all that important in the larger scheme. Would you fire a programmer who let a bug slip through? Would it be acceptable to give a zero to a student who had a bug in their code? I've seen successful books with spelling errors. They should be minimized, but having a single spelling error doesn't mean the write gets 0 income. I've seen math professors forget a negative many times, doesn't mean they are idiots who don't know math or that they are no longer worth learning from. I personally find professors who are overly strict to be compensating for not being able to make their material appropriately difficult. Effectively, if they didn't do this, their class averages would be abnormally high. I have many professors with hard classes that taught well yet didn't need to resort to this style of grading to think that one must do this to avoid grade inflation.

ryandrake 2y ago
> Would you fire a programmer who let a bug slip through? Would it be acceptable to give a zero to a student who had a bug in their code? I've seen successful books with spelling errors. They should be minimized, but having a single spelling error doesn't mean the write gets 0 income. I think it depends on the error. For the code example, the code should at least be compilable, with no syntax errors. How could you give anything other than an F on a code assignment that doesn't even compile?
- WalterBright 2y ago
  Caltech had other institute rules: 1. Calculating intermediate results and carrying them forward was an automatic F. (This was because accumulating roundoff errors was a big no-no, and often if you worked through the algebra a lot of the terms that contributed to errors would cancel out.) 2. Coming up with absurd answers got you not just an F, but negative credit. Caltech expected students to know what was reasonable. Negative energy was not reasonable. If you produced such an answer, you'd need to annotate it with "I know this answer is absurdly wrong, but I cannot find my mistake" and then you'd get partial credit for how far you'd gotten in the derivation. Caltech was training engineers and scientists, where such mistakes can have disastrous consequences.
- SkyBelow 2y ago
  Depends upon why the code doesn't compile. It is a red flag, and if the student really didn't test it or try to compile it at all that would be a major problem. But even being able to compile code isn't must do in the business world. For example, just today I was fixing a CI/CD pipeline issue where local builds worked fine but the pipeline didn't. I don't think whomever caused the underlying root issue deserves to get fired for that. There are obviously mistakes that go too far, but these generally involve an element of malicious intent. If a junior takes down prod, that's probably a sign that some security settings are missing. Only if they knew they were taking down prod, knew they shouldn't, and did it anyway should something like firing be on the table.

ryandrake 2y ago

> Would you fire a programmer who let a bug slip through? Would it be acceptable to give a zero to a student who had a bug in their code? I've seen successful books with spelling errors. They should be minimized, but having a single spelling error doesn't mean the write gets 0 income. I think it depends on the error. For the code example, the code should at least be compilable, with no syntax errors. How could you give anything other than an F on a code assignment that doesn't even compile?

WalterBright 2y ago
Caltech had other institute rules: 1. Calculating intermediate results and carrying them forward was an automatic F. (This was because accumulating roundoff errors was a big no-no, and often if you worked through the algebra a lot of the terms that contributed to errors would cancel out.) 2. Coming up with absurd answers got you not just an F, but negative credit. Caltech expected students to know what was reasonable. Negative energy was not reasonable. If you produced such an answer, you'd need to annotate it with "I know this answer is absurdly wrong, but I cannot find my mistake" and then you'd get partial credit for how far you'd gotten in the derivation. Caltech was training engineers and scientists, where such mistakes can have disastrous consequences.
SkyBelow 2y ago
Depends upon why the code doesn't compile. It is a red flag, and if the student really didn't test it or try to compile it at all that would be a major problem. But even being able to compile code isn't must do in the business world. For example, just today I was fixing a CI/CD pipeline issue where local builds worked fine but the pipeline didn't. I don't think whomever caused the underlying root issue deserves to get fired for that. There are obviously mistakes that go too far, but these generally involve an element of malicious intent. If a junior takes down prod, that's probably a sign that some security settings are missing. Only if they knew they were taking down prod, knew they shouldn't, and did it anyway should something like firing be on the table.