11
\$\begingroup\$

The current de-facto policy we have regarding homework-like questions is to require an attempt at solving the problem by the OP. While this keeps the most blatant cases of laziness out of our question list, there are at least two kinds of questions which I personally don't like that slip through:

  • "check my calculations" - this is when the OP asks yet another question about equivalent resistance or Kirchhoff's laws. They made their calculations, but suspect there's a mistake and ask to double-check it. This is a waste of time IMO, they could simply present their solution to the professor giving them the course, or simply enter the circuit in a simulator and get the right answer.

  • "monkey with a typewriter" - this is when the OP doesn't know how to solve the problem, but since questions without an attempted solutions are closed, they will present a nonsensical solution to fulfill the rules. A proper answer will still have to start from scratch.

Yet, the requirement to present an attempted solution keeps away some legitimate questions where the OP doesn't have an idea where to start. As an example, if the OP doesn't know about delta-star transformation, they will be completely stuck with a problem where they need to apply one. They can of course resort to the "monkey with a typewriter" strategy, but I don't think that's very educational.

Perhaps a better policy for homework-like questions would be to favor conceptual questions, which don't ask for specific numbers. Such a policy would keep people who are too lazy to plug their numbers in well-known formulas away, and also get rid of "check my calculations" questions. Additionally, we would be able to keep questions which the OP doesn't know how to tackle, without giving them an incentive to present a nonsensical solution as an excuse. And conceptual questions make better duplicate targets too, because exact numbers don't matter.

Now, I'm not arguing that engineering questions should not include numbers. In fact, if the OP is debugging a circuit and had measured a voltage at some point which they think is abnormal, the exact value of the voltage is crucial for answering. But I believe it's fairly easy to distinguish homework-like questions from engineering ones: they feature circuits with no practical application, have artificial constrains (e.g. a requirement to calculate something that could be easily measured) and so on, so I don't think we'll be throwing out the baby with the water.

I'm also not opposing to homework-like questions which include numbers as an illustration. If the OP have an question with numeric values printed on the schematic, they don't have to remove them. What would be forbidden is to ask for a specific number: be it the equivalent resistance, voltage between points A and B, or the cutoff frequency.

Also note that we don't have to drop the "attempt at solving" requirement, though I would suggest we lax it quite a bit. That is, the OP who has been tasked with equivalent resistance calculation and didn't try anything will still get their question closed, either because of "no attempt" rule, or because they asked for a specific numerical solution. The OP who worked on their problem and got stuck in the middle will be able to both demonstrate a reasonable attempt at solving and ask about a specific step rather than the final answer.

Any thoughts on the above? Examples where such a policy would be lousy?

Edit: please don't hesitate to downvote if you dislike the idea. So far all answers I got are critical, yet there's only one downvote.

\$\endgroup\$
  • 3
    \$\begingroup\$ I'm not sure we should get rid of "check my calculations" questions entirely. I've seen more than a few questions of the form "I got this result, but it doesn't make sense, where did I make a mistake in the calculations", and I've had more than a few unhelpful professors in my time at university who wouldn't be of much help for that kind of question. \$\endgroup\$ – Hearth Oct 15 '18 at 18:07
  • \$\begingroup\$ @Felthry Well I've seen several questions where (1) calculations could be easily checked by a simulator and (2) the OP didn't have any hints as to why their calculations might be incorrect. In the end, if the question is about which law/formula to apply, it's an EE question. If on the other hand all formulae are known and a numerical solution is needed, it's a math question. \$\endgroup\$ – Dmitry Grigoryev Oct 16 '18 at 6:38
  • 1
    \$\begingroup\$ Regarding your recent edit, it's a little late now but maybe it would have been better to split your post into a Question+Answer. I upvoted the question because it's a good question that brings up an interesting topic that we should discuss but I'm not sure that I like your suggested solution. \$\endgroup\$ – pipe Oct 19 '18 at 7:55
  • \$\begingroup\$ @pipe Yeah, perhaps that would've been better. Anyhow, my takeaway from this discussion is that most people seem pretty happy with current rules and don't mind calculation questions, so I disgress. \$\endgroup\$ – Dmitry Grigoryev Oct 24 '18 at 13:53
3
\$\begingroup\$

Ethically they need to have an attempt at a solution and ethically we need to require an attempt at a solution. Outside if that not much matters, if it is conceptual, then invite them to write down what the solution to the problem is. There are too many people that come here thinking that they can shave a few minutes off their homework by posting the questions (sometimes verbatim) and not even thinking about it. This does a huge disservice for them because they aren't building the neural circuits necessary to become good engineers.

Check my calculation questions are fine, as long as they provide a calculation.

\$\endgroup\$
  • \$\begingroup\$ First, I specifically stated that the requirement of an attempt doesn't have to be dropped entirely, only when it doesn't make sense (the OP is lacking a crucial concept which prevents them from producing any meaningful solution), because coming up with a bogus solution to fulfill a requirement is not very smart. And second, I don't remember the last time I learned something useful from a question or answer which revolves around calculations. \$\endgroup\$ – Dmitry Grigoryev Oct 16 '18 at 6:41
  • 2
    \$\begingroup\$ I disagree with "check my calculations". As Dmitry writes, just put them into a simulator. A question that can be answered with "That's correct." is not suitable for the Stack Exchange format IMO. \$\endgroup\$ – pipe Oct 16 '18 at 7:46
  • 3
    \$\begingroup\$ @pipe Although I understand Dmitry worries, the requirement to be able to use a simulator is not something that comes first in many uni courses even now (at least in Italy). Some students may have legitimate difficulties in carrying out the calculations with pencil and paper and still not know that simulators exist. I almost bet there are still some "old-school" professors that shun students that rely on simulators for verifying the correctness of their calculations ("never trust the PC over your brain"), even if the calculations imply a 5x5 matrix inversion! \$\endgroup\$ – Lorenzo Donati supports Monica Oct 16 '18 at 17:24
  • \$\begingroup\$ @pipe I mean, It is hard to say it is fair to rule out any "check my calcs" questions (unless it is just a "plug numbers in formulas" question). Some may be legit, other may show laziness or lack of study. \$\endgroup\$ – Lorenzo Donati supports Monica Oct 16 '18 at 17:26
  • 1
    \$\begingroup\$ @LorenzoDonati Those could be closed as duplicates of a canonical "How do I check my calculations in a simulator?" question though, at least for DC. \$\endgroup\$ – pipe Oct 17 '18 at 8:14
  • \$\begingroup\$ @LorenzoDonati The fact that there are professors which want students to invert a 5x5 matrix doesn't imply that we have to do that for them, and Math SE is better suited for that kind of questions anyway. Also, learning how to use a simple simulator such as this one doesn't require a uni course. \$\endgroup\$ – Dmitry Grigoryev Oct 18 '18 at 8:08
  • \$\begingroup\$ I'm not sure how ethical it is to require prior effort from questions deemed to be homework, yet accept questions which show no effort at all (e.g. "what is voltage?" or "explain how this circuit works") \$\endgroup\$ – Dmitry Grigoryev Oct 18 '18 at 10:45
  • \$\begingroup\$ @DmitryGrigoryev it is pretty clear (to me anyway) which questions are design related and which are academic. Most academic questions have little relevance to the real world and are designed specifically to test a concept, this makes them easy to spot. \$\endgroup\$ – Voltage Spike Oct 18 '18 at 15:40
  • \$\begingroup\$ The worst thing is I have seen several times people posting questions directly from an exam or quiz and asking for a quick solution. \$\endgroup\$ – Voltage Spike Oct 18 '18 at 15:40
3
\$\begingroup\$

For many years, the SE policy has been that we don't care why someone is asking a question, but only about the quality of the question itself. That is, a student asking for help with their homework will receive the same treatment as an electrical engineer looking to verify their design. In either case, they have to demonstrate an insight in the topic and their attempts to solve the problem so far.

A question which provides a complete calculation of some electronics problem, with provided schematics etc, shows sufficient research effort. It is fine and on-topic.

As for the "monkey with typewriter" case, the attempt provided has to be relevant to the question. These attempts are often easy to spot. If the "research effort" or "attempted solution" is just fluff and not relevant to the question, it should be treated like a question with no effort at all: close as too broad.

\$\endgroup\$
  • 1
    \$\begingroup\$ There's a difference between engineering and homework questions which cannot be ignored. For example, "equivalent resistance" questions have a simple engineering answer ($10 multimeter) which is completely useless if the question is homework. The fact that we don't care why a question is asked doesn't mean we have to accept non-engineering questions on an engineering site. \$\endgroup\$ – Dmitry Grigoryev Oct 18 '18 at 8:03
  • \$\begingroup\$ @DmitryGrigoryev Each question does of course need to be handled on case-by-case basis. A question about trouble-shooting a circuit, where the OP doesn't even have access to a multimeter, smells like a candidate for closing. \$\endgroup\$ – Lundin Oct 18 '18 at 8:14
  • 1
    \$\begingroup\$ That's not what I meant. Homework questions often come with artificial constraints: cannot measure X, solve without Fourier transform, don't use simulator, etc. which make them distinct from engineering questions. So I fully agree that we shouldn't care why a question is asked, but we still may decide that some non-engineering questions are off-topic. Even if asked by Olin :) \$\endgroup\$ – Dmitry Grigoryev Oct 18 '18 at 8:21
  • \$\begingroup\$ @DmitryGrigoryev Well, that's a different matter not addressed by the question: should we allow questions with artificial/academic requirements? Over at meta SO, there have been suggestions to ban such questions, but they were shot down in flames. Maybe we should create a separate thread about it for meta EE. \$\endgroup\$ – Lundin Oct 18 '18 at 8:28
  • \$\begingroup\$ I'm not against such questions in general, just against those which boil down to number crunching. If the OP ask us to calculate something, the question is out. If the OP calculated something and asks to double-check, that's out as well. But if the OP found out their answer is incorrect (via simulator, answers in the workbook, whatever) and wonders why their approach didn't yield the expected result, that's legit, since it's a question about the approach, not about numbers. \$\endgroup\$ – Dmitry Grigoryev Oct 18 '18 at 8:38
  • 1
    \$\begingroup\$ This is a good point, and artificial constraints shows up in other questions too. I personally tend to skip questions saying "I don't want to use a microcontroller" because it's no longer really engineering, it's a mind puzzle or a history lesson. \$\endgroup\$ – pipe Oct 18 '18 at 9:07
  • 1
    \$\begingroup\$ I find a worrying trend here, which is that many of the responders seem to believe that simulators are an infallible source of correct answers. After a lifetime as a working engineer, I'm more inclined to take the view that "90% of all simulation is garbage" (a quote from a keynote speech at a conference I attended, back in the 1970s - but some basic principles of engineering are timeless!) \$\endgroup\$ – alephzero Oct 23 '18 at 21:00
  • \$\begingroup\$ @alephzero I had basic stuff in mind where simulators are pretty accurate. E.g. Kirchhoff laws in a passive circuit, equivalent resistance, transient analysis of basic RC or RLC, ideal opamps etc. \$\endgroup\$ – Dmitry Grigoryev Oct 24 '18 at 12:40

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .