Hacker News Reader

Comment by commandlinefan | Hacker News Reader

> every word of the assignment creatively misinterpreted

OTOH when I took operating systems I got an assignment that said “implement a job scheduler, using FIFO, LIFO or round-robin job scheduling”. So I picked FIFO, got it working and I had time left over so I thought, “what the hell? I’ll do LIFO too”. So I did, and I still had time so I took a crack at round-robin, but I didn’t have time so I turned it what I had, proud of myself for going above and beyond.

I got back a 66 on the assignment. I asked why and he said, “you didn’t even attempt round robin”. I pulled up the assignment where it VERY CLEARLY said “or” and he said, “well, it should have been obvious I meant ‘and’”.

View on HN

> every word of the assignment creatively misinterpreted

I hear you, communication is hard, but trust me your statement and the authors are an AND not an OR :) the difference is scale. the students are about 1v30 when it comes to opportunities for badly written assignments. In a typical semester, I am 1v120+ hostile attacks on my communication skills

The number of students I have that actively presume I am trying to trick them is truly astonishing. It's a learned behavior, sure, but it is the behavior.

I've put parts of answers in the question text, underlined them, and then had students tell me they thought it was a trick. I'm at the point when I write tests I leave pretty blatant hints because I have found they are a useful way to punish students who have opted out of thinking and try and brute force answers. I've had students actively invent parts of assignments in their head. In the last one, I asked for three examples in a short answer problem...and gave one of them as a 'for example'. less than 10% of people used that. More than 10% told me they thought it was a trick.

I'm not mad, I'm worried. I get why IT is moving towards zero trust, but it is fundamentally at odds with actual learning. The saving grace is that I can actually use the term 'headcannon' with this generation and they know what I mean.

> The number of students I have that actively presume I am trying to trick them is truly astonishing.

> I've put parts of answers in the question text, underlined them, and then had students tell me they thought it was a trick.

Why wouldn't they assume that?

Tests are nothing but games where students figure out what answers you want. They can do that by learning the material, but there's a metagame. What were you thinking when you wrote down this question on the exam?

For example, once I took a english test which asked me the correct pronoun for ships. Choices: he, she, it, they. I know that "she" is almost always used to refer to ships, but I also know that "it" is also grammatically correct. So which is it? Is the teacher trying to test my understanding of this historical tradition? Does the teacher think "she" is antiquated? Will I lose if I pick "she"? There's enough uncertainty here for me to pause even though it's an absurdly simple question. I know the answer but I'm not 100% sure what will happen if I choose it.

There are serious consequences to getting these answers wrong and students simply can't afford to trust you.

Entire books could be written about this teacher-student metagame. Another example: choice order. Suppose the choices to the above english question were ordered so:

  Which pronoun is used to refer to ships?

  A. It
  B. He
  C. They
  D. She

Deliberately positioning an almost correct answer as the first choice was extremely common in the schools I attended. A student who's overly anxious or running low on time might read that, pick choice A without even reading the other options and move on. The second we noticed this, "letter A is never right" became a meme and to this day I feel uncomfortable choosing the first option presented to me in any context. Preparatory schools actually offered test taking classes where we were taught to read questions backwards because of this.

Teachers eventually noticed that we noticed and suddenly the right answers were the first choices. Do you have enough confidence in your knowledge to assert that letter A is correct? Now it's a deeper level metagame: does this teacher know that I know?

This is what you get when education determines your future job prospects. It is no longer about "actual learning". It is a game and our future hangs in the balance.

I actually would have chosen "they" on that multiple-choice question, because it asks about ships, plural.

... And you would have been right.

I think we are on the same page.

My point wasn't that they shouldn't or should, its that they do. As you note it has been socialized into them. The problem is systematic, not something any one educator can fix. Bad teachers, and bad educational experiences create long term damage beyond not learning content.

Good tests actually assess learning - what you mentioned is not a good test. Just like a test that causes stress ends up measuring stress isn't measuring learning.

> The number of students I have that actively presume I am trying to trick them is truly astonishing.

Because we all have experience of you guys doing just that.

I once had a question on an exam for high voltage power “why do high voltage transmission lines have 3 wires”. The answer I put seemed obvious, that each wire is for one of the 3 phases of electricity. This is the standard config for HV lines in the US.

The answer was wrong, because the professor was talking about INSIDE the cable and wanted an answer relating to how the electromagnetic field rotates around the wire so they bunch wires inside each line to improve efficiency.

Professors, especially elite ones, have seemingly perfected the art of being vague. As a student you have zero recourse against them for their terrible communication skills, your only hope is to dodge the whacky ones by befriending upperclassmen/being in a frat.

> In the last one, I asked for three examples in a short answer problem...and gave one of them as a 'for example'. less than 10% of people used that. More than 10% told me they thought it was a trick.

What did you expect? To me, it would be painfully obvious that the already given example is void and no longer up for grabs. I would never construct such an exam question. Dangerous slope, but in this case obvious.

I have to say I think it's quite possible and comical that the professor interpreted your work on two of the three questions as additional evidence that you did know it should be an "and", since I would guess relatively few students would attempt any more than _one_ permutation if the assignment only asked for one.

Damned if you do, damned if you don't.

But what did the other students the class do? If almost everyone in the class did only one, the professor would have to give almost the entire class 33% for a mistake he made.

Looking back, I should have asked around. The job scheduler was one of several options so not everybody picked it. Still, somebody else must have.

The professor actually did relent and gave me a good (not perfect, but good) grade. Then he judged me extremely harshly on the final exam so I ended up with a B in the course anyway.

That’s pretty sad. As a CS Professor, he should understand the distinction between a conjunction and disjunction.

I think It’s a question of scope not of the logical operator. One way to interpret the question is:

1. You choose X or Y or Z

2. You provide a scheduler that does the thing you chose

Another is:

1. You provide a scheduler where I choose X or Y or Z and the scheduler does the thing I chose.

The question is whether the disjunction applies to the input configuration of the scheduler or the output of the assignment. That is, whether the output of the assignment has type (X_scheduler | Y_scheduler | Z_scheduler) or type (X | Y | Z) -> scheduler.

I agree with you, but not as the assignment is worded:

> “implement a job scheduler, using FIFO, LIFO or round-robin job scheduling”

To me, "using" modifies "implement"; that is, the teacher is telling me to implement something, and then telling me what to use. (That is, "using" does not modify "job scheduler".) If the teacher wanted the second interpretation, they should have said:

"Implement a job scheduler that can use FIFO, LIFO, or round-robin scheduling."

In this case, "that can use" is clearly referring to the job scheduler, not to me. To be fair, though, I think some students might still (reasonably!) believe this means they only have to implement one algorithm. I would probably use this second wording, but also change the "or" to "and". Even better, though more wordy:

"Implement a job scheduler where the user can select between FIFO, LIFO, and round-robin scheduling."

Let's also remember that boolean logic and English usage are not the same thing. I don't think my two suggested wordings would be considered "over-specifying" in the same way that OP is talking about. They're just a better use of language.

The teacher here should just have re-read the problem instructions, shown understanding and empathy toward the student, and either a) just changed the grade right then and there, or b) offered to let the student complete the rest of the assignment and turn it in again. And then made the wording of the problem more clear for the next batch of students.

Not more clear, he should've fixed his mistake. Or is not and. Period. I agree with you, but I wanted to emphasize the mistake

As an educated human, he should understand that when you provide written instructions it is implicitly a "do what I say" and not "do what [you infer] I mean".

It is obviously unfair and unprofessional to penalize the student for the professor's error.

Tell that to all the math professors that go: "this part is obvious"...

While I think a student could read the instructions in the way the teacher intended (though I would not be one of those students either), I think the problem here is that the teacher is a poor communicator, and is too arrogant to believe that they could be fallible here.

The fix for this particular issue isn't over-specification, it's changing one word in the instructions. Or at most, adding a few more words to make things clearer.

Reply on news.ycombinator.com