Comment by jrh206 | Hacker News Reader

Comment by jrh206 | Hacker News Reader

Parent Full thread

I don't think this is a good parable (regardless of veracity).

Occasionally students get hung up on difficult questions, determined to solve them, losing time for the remaining questions. Yes, this is an anti-pattern, but it is a fact of life. By incorporating an impossible question, you waste the time of honest[1] students.

Some will recognise that the question is impossible and move on. Some with good exam technique will know not to get stuck on a problem they can't solve until they've had a go at all of the other questions.

But some honest and hard-working students will have got a worse grade than they deserved because of this stunt.

[1] Consider, as well, 'desperate' students - students who need a certain grade for some reason or another. Maybe this class isn't these students' priority, but they need to pass to be able to study the classes they love. An impossible question could drive these students to cheat, when they never would consider it otherwise[2][3].

[2] Consider the detail "about half the class left the room to use the bathroom during the test - far more than the usual".

[3] This bears resemblance to vigilante groups targeting peadophiles. I don't know what to make of that.

Wouldn’t you say that not getting hung up on the little details is a vital real-world skill? Like you say, it’s a fact of life that some do. Should we ignore it then, so that everybody has a “fairer” chance?

In fact, this is one of the most exam-tackling strategies: spot and rake up points with low-hanging fruits throughout the test.

People make a lot of assumptions when taking a test, and a common one is: when I answer all questions correctly, I get a the best score (a 10, an A+, or a 1, depending on your country's system). So they try. They answer in "test mode". My guess is it would affect mediocre students most.

But, to put it in your words: exam tackling strategies are not vital real-world skills.

Since when are tests about examining vital real world skills? If tests were about that, reference lookup, team work and phones would not be against the rules. You can and should always do that in the real world.

The truth is tests are merely a game between student and professor. A game born of arrogance and gatekeeping. There are even metagames being played as exemplified by this article: would the professor really ask such as question?

I dont think general engineering exam should contain traps to test your ability to "not get hung up on little details" in the form of unsolvable exercise needed for full credit.

A bit offtopic to article, engineering exams tend to be about noticing little details and not ignoring them.

Are those then “little details” or “traps”?

The story doesn’t report that only an elite few were able to solve a specific next-level question to get full marks. It says some people are so hung up on grades that they went straight for the honey pot.

The article talks about whole unsolvable exercise needed for full credit, which is a trap. It is not little detail at all. You used the expression "little details", I have put it in quotes because I dont think it is accurate characterization at all.

> It says some people are so hung up on grades

So? Practically, grades matter pretty often. And even if they did not, there is nothing wrong with being motivated by grades or having grades related ambitions.

> The article talks about whole unsolvable exercise needed for full credit, which is a trap.

You say this like full credit is required to pass. A perfect score is often practically impossible, and at least a big achievement in most of higher education, not something most people can walk in expecting to get.

> You say this like full credit is required to pass

I don't. I did not said nor implied that. I said full credit, because I meant full credit. I did not meant "required for barely passing".

> A perfect score is often practically impossible, and at least a big achievement in most of higher education, not something most people can walk in expecting to get.

It is pretty often practically possible and achieved. In any case, motivated talented hard working student is entitled to expect full credit to be possible.

Average student is entitled to expect all exercises to be possible to solve and within range of what was taught in the course. Even if he personally cant solve everything, it should be doable.

The story was admitted fake by the author, so the details of the story aren't to be used to justifying anything.

I would argue that students who haven't learned this skill have a certain expected value of penalty already applied, when they take a legitimate test. An impossible question compounds the penalty.

EDIT: In other words, if you know to hoover up the easy marks, you get a better score anyway.

It breaks trust with the teacher. You are paying for knowledge and an honest evaluation of your mastery of that knowledge. That real-world skill is not a good fit for a test without the foreknowledge or training.

An alternate reading ...

If a teacher takes no real measures to prevent some students from cheating, that breaks trust with the honest students. This story - true or not - is of a teacher trying to create trust with the honest students by ensuring that at least some of those who cheat don't get away with it.

What is cheating anyway? Looking up the answers in one's books or phone? Talking to your fellow students? That sort of artificial difficulty shouldn't even be part of the rules in the first place. If your tests are so easy they can be solved via Google, they're pointless and it doesn't really matter if someone is cheating in the first place.

Honestly the problem is this mass education model where professors have hundreds of students to assess. Tests are necessary in this environment because they scale to any number of students. If people want an honest appraisal of their skills, a mentor who focuses on a single student exclusively is much better positioned to provide it.

> What is cheating anyway?

Not abiding by the rules that have been set.

> Looking up the answers in one's books or phone?

Yes, if the rules have said that you are not permitted to do so.

> Talking to your fellow students?

Yes, if the rules have said that you are not permitted to do so.

> That sort of artificial difficulty shouldn't even be part of the rules in the first place.

If you are studying a subject that is intended to have you gain certain skills, then it is reasonable and appropriate to assess whether or not you have done so. Being able to recall certain facts and to carry out certain procedures without external assistance is sometimes/often included in the skills you are expected to acquire as part of the course.

> If your tests are so easy they can be solved via Google, they're pointless and it doesn't really matter if someone is cheating in the first place.

Suppose you are learning calculus, and your instructor wants to know if you can differentiate a particular function from first principles. This is an important skill to ensure you are ready to move to the next stage, but it's trivial to search for it on the internet.

This is just one example that counters your assertion.

> If people want an honest appraisal of their skills, a mentor who focuses on a single student exclusively is much better positioned to provide it.

True. Given the current trends in teacher numbers versus student numbers, that's unlikely to happen. So in the meantime we must do what we can with mass assessment, and try to make that assessment fair. Preventing cheating is part of that.

Searching for help on the internet is also a valuable real world skill. Does that not mean then that the cheaters should receive extra credit instead of a zero?

Not for proctored exams that are designed around you studying in the lead up to the exam, not looking stuff up during the exam.

If it was an open-book exam then arguments could be made that looking stuff up online is an extension of looking it up in the official textbooks (unless you are _only_ allowed the official textbooks... but you get my drift).

Every school and uni exam (for the most part) will have past papers you can test yourself with. The exam doesn't deviate a whole lot (in my experience) from the past papers with the changes usually revolving around the made-up scenario, e.g. "a boat with a speed of 34kph..." vs "a plane travelling at a speed of..."

They could have studied the past papers and done well but they wanted the easier route.

So, no, they cheated. They deserved their punishment.

Not to mention that this was posted and answered online, and it's possible that one or more of the students might have actually seen the question beforehand and memorized it. Some people have eidetic memories.

I have no respect for cheaters, but catching innocents in this little trap is absolutely unconscionable to me.

This is mitigated somewhat by "there was one particular question that wasn't related to what they had all been taught in class"

But I agree

I don't think that actually is a mitigation. I had professors that asked questions on a take home that were related but had not been discussed during any class. I still find some of the views on this strategy to be odd.

It's incredibly unlikely that you wild stumble across a specific random false trivia fact in the whole of the internet, and internalize it without noticing it's wrong.

Not to mention, many professors set questions of this sort that you need to correctly answer for an A.

Reply on news.ycombinator.com