-6

You all know the problem, trying to eliminate the bad homework stuff and keep the good.

I code off and on and my personal website has a filter on every input box to eliminate the stuff I don't want to see on the comments, the usual limitations...language, abuse, etc.

My suggestion is as follows (I can't find a duplicate of it. I am only here 10 days so please be nice in your comments and replies. Thanks.):

Kids come onto the sites from any page so target the question boxes.

Install a filter that spots obvious phrases like "My teacher says I have to have this...", etc.

When these phrases are spotted, produce a popup that stresses the homework policy again and scares them (a little) with it. Make it more kid-oriented and suggest a list of kid-based sites that they could go to.

Stop the "Can you do this straight away part" by telling them that the question will be reviewed for quality and there will be a one week delay period before an answer.

Then, after the question is written, analyse the wording of the question for the common bad homework phrases using the filter and if they occur then shunt those questions into a review area, so that it's NOT immediately added to the list of questions, and it really does have to wait a week.

So for example, if the phrase "in your opinion" came up in a physics question, then it would be flagged straightaway.

If it is a bad question, then send out the standard line about "it will not be answered" and then delete it.

This above part seems easy (to me) to do with a few bits of whatever code integrates with the underlying Stack Exchange code). Is it that easy, you tell me, I don't know the Stack Exchange code foundations.

Now you want to keep the good questions. The more advanced kids will know to stick to their guns and stay here and they are the ones you WANT to stay here. Maths.SE must have the same (if not worse) problem and between the population of physics and maths users there must be enough talent to produce all sorts of correlation code between phrases used in good questions. The data exists and the system can be tweaked as more homework stuff comes in.

Funnel the good homework stuff for manual review and then release them to the users.

It can't cater for all kids, but what's the current rate of unanswered kids' questions now anyway?

That's it. My question is (because I havn't enough experience of the site yet), is it doable?

  • 10
    The people that ask bad questions about their homework are not going to be stopped in the least by you trying to tell them more an more frequently/harshly that they might be doing something wrong. They aren't even going to be reading the messages, just clicking through them. – Servy Mar 16 '15 at 19:23
  • 4
    wonder how this could help folks who already ignore 15 (fifteen) warnings shoot straight in their face. Those who "...should maybe get electrocuted if they touch their keyboard and then they still post on meta innocently asking what possibly could be wrong..." – gnat Mar 16 '15 at 19:38
  • 2
    @gnat or they'll dub the shocks a feature – Gilles Mar 16 '15 at 19:42
  • @servy do me a favour, i left out a paragraph ...please read it again and see does it make more sense ... maybe i have made the idea a bit more doable – irish physics Mar 16 '15 at 19:43
  • 3
    @irishphysics The phrases you've used as examples can most certainly be in entirely appropriate questions about homework. Deleting all questions with those phrases and just assuming that they're all bad is going to have a very high false positive rate. It'll also have a high false negative rate as so many of the worst questions that are based on homework won't have any of those phases. – Servy Mar 16 '15 at 19:45
  • Bad is subjective which means it takes someone to review it. There's already a good system in place for new users (the First Posts queue, similar ones exist on all SE sites) and most good users are well-aware of the ability to flag as low-quality. Seems that you just want to automate a system that, from my experience, works just fine. – Kyle Kanos Mar 16 '15 at 19:45
  • @KyleKanos they say, The “First Posts” review queue is ineffective and toothless – gnat Mar 16 '15 at 19:47
  • @gnat: Perhaps that is the case with SO, but on the smaller sites, I'd disagree (as do some other people). – Kyle Kanos Mar 16 '15 at 19:51
  • @servy thanks very much for that..... it might fall down because we can't discriminate enough, ... you have more experience than me.... if the phrase "in your opinion" came up in the physics area that would immediately flag it . A different set of filters for each section of the entire site? – irish physics Mar 16 '15 at 19:58
  • 1
    I'm not going to say that what you're asking for isn't doable. I'd bet a large sum that a dedicated team could implement this well. But I'd also bet that the effort of doing this is far more than the result is worth. The current system of relying on users to review it works adequately and allows users to use their personal judgement in difficult cases. This might be doable and might improve efficiency, but there's no real problem that this fixes nor pressing need to implement it. The benefit probably is not worth the effort – Jim Mar 16 '15 at 20:42
  • @Jimnosperm thanks for your considered response ...i'm hindered by the fact I don't much about scale /structure of SE ....biggger than my blog tho, that's for sure! regards – irish physics Mar 16 '15 at 20:53
3

I understand your passion to resolve a problem, or even alleviate it a bit, and this is of course always welcomed.

But sometimes a problem cannot be captured as it is happening, or it is not worth the resources to do so.
Often, problems simply require a post fix - in this case community moderation - and as we have a strong community moderation system already setup, it's probably much more efficient to utilise that.

(Note, in this answer I state "frequency of problem" not "frequency of homework questions" - some are ok)

Not worthwhile

Install a filter that spots obvious phrases like "My teacher says I have to have this...." etc.

Given:

  1. W% of users won't post any phrase at all
  2. X% of users will ignore any/all warnings
  3. Y% potential false positives catered for (to avoid wrongly moderating users)
  4. Z% frequency of the problem this attempts to resolve

I just don't see this being worthwhile as we'd have a complex set of algorithms to try to catch the "odd" question here and there.

Warn them, tell them, tell them again, plead with them...

When these phrases are spotted, produce a popup that stresses the HW policy again

How many levels deep does one go before we give up and say "this user is never going to adhere to the rules, read the rules, or care"?

They are warned, the rules are clear and readily available.

If they got that far, and post what you consider a "bad homework question", then they will be told in comments.
If they do not edit/fix their question from warnings in comments, then this user is unable to be cleansed of their demons.

You could show them 10 warnings, ranging from "Please change this question" to "pretty please" to "We know where you live...".
You just can't tell, persuade, or educate some people. They simply cannot be bothered/don't care.

You give them the info, once, and if they choose to ignore it, then they lose a potential fantastic answer from a professional.
We're here to advise, help where help will be taken, but we're not here to hold everyone's hand who is not willing to put in some pre effort or have a little humility and resolve an issue they are post made aware of.

Either way, I think the current path and entire process is adequate.
We already have community moderation, and while these problems are annoying, and a solution would be great, it's only closing a few bad homework questions and this proposal is too "meaty".

One week answer delay

Stop the "Can you do this straight away part" by telling them that the question will be reviewed for quality and there will be a one week delay period before an answer.

No idea what this aims to resolve or help with.
Maybe I've misunderstood it. (?)

As I understand it:
With a "one week answer delay", where does the question "wait"? In that time it'll be pushed so far down the list, page X depending on the tags - i.e. PHP tag probs > 1000 pages. So they'll never get an answer anyway.

Also, this is not how Stack works. It's (almost) as simple as:

  • If the question is not acceptable, close it
  • If the question is acceptable, answer it

Use of resources

This will be a lot of time spent discussing, implementing, testing, maintaining and most certainly tweaking given the likely false positives that will arise.
It won't be easy to roll out a perfect system, and also given the plethora of potential word combinations where some are from the bad list and some not, not every scenario can be pre established, and not every false positive can initially be accounted for.

It really is a potential minefield.
If we were getting 10 "bad" homework questions "per hour" then perhaps this might be looked at.
But even then it would likely be something much more simple than this proposal, just a quick and dirty way to "slow them down a bit".

Alternative use of resources

I'd much rather we had some new systems and time spent on making some of the "general" bad questions go away. The homework questions can come through thick and fast after term holidays end etc, but they are not exactly a serious issue.

And I'd rather forgive a teenager for being, well, a teenager*, and help them if they say "sorry I'll edit my question" than the lazy bone idle people who post poor, lazy, no code, questions, and ignoring the comments of people telling them as such, as they know someone will answer them anyway.

* I mean no insult - I have an 18 YO step son doing Engineering degree/A levels, and know how hard it is, often a lot of info, and sometimes they are desperate for some help they don't think about pleasantries.
This does not excuse it, but I give "kids" as you put it a little more leeway than "adults" who should be more mature.

  • 1
    thanks for your considered and helpful response, my full reply is below your post – irish physics Mar 17 '15 at 1:04
4

Firstly, thanks for your time in sending me the above. Believe it or not, we are at one on this, despite the way I composed my question. I edited my post based on some of the comments I got and some more reading on the subject. The first draft was much more student friendly, I think.

Just hear me out for a minute.

I'm a complete self study person in quantum physics (no university contact at all) who has learned more about how real science works (and some science people really behave towards each other!) in the last 10 days from this site than I ever imagined.

Think Robinson Crusoe teleported to NYC, and you will get the idea

I do 100% agree with your last paragraph. I have given lots of tutorials, and I deal with the online student exactly the same way as at my home, trying to get them to keep asking me questions and then think it through themselves as far as they could until the next hint.

The reason I posted the question was because I was absolutely furious at the way one particular student was treated by a "teacher" who either could not care less about the student or was too tired or, I don't know basically had a bad day and should have ignored the post if they could not handle it. I know, I should have flagged it, but I didn't.

That student is now going to stay off physics, for a while anyway, because he/she was shouted at, basically. It made me mad, because I could see he/she had potential.

Then I saw lots of posts about "too many bad homework questions", "should we ban homework", etc., and I thought: if the students are getting bad answers partially because of too many bad questions stressing the "teachers" out, could I think of a way of removing that stress by shunting the bad questions away from them in the way described in my post and then see what people with more experience than me had to say.

OK, it's not doable (the Stack Exchange site is way too complex and way too big, but I didn't realise that until this post), and it's time to move on...there is no easy way to avoid bad questions preventing the "teachers" from occasionally dealing badly with the students (or from some students behaving badly). It's just that I never in my tutorials treated the students in any way, but the best way I could. You really would go a long way to help a student that shows academic promise; it could change their whole life. I'm telling you what you know already here, sorry!

To wind up, the existing scheme on Stack Exchange is the best we have. Most students get treated with patience, some students just want the easiest way out and don't get the long term picture of understanding being the goal, not just rote, and unfortunately some just get unlucky with their "teacher".

And, worst of all IMO, some students submitting homework in a badly presented way because of real pressure from their parents that they will be a failure if they don't win the Nobel prize. Pretty sure you have come across this in some way yourself.

Sorry about the "kids" rather than students reference, but I have done enough tutorials by now (I'm 54) that it's just a habit that I carried from my real life tutorials, answering the OP, especially if it's a basic question, as well, a kid!

That's basically it, question answered and closed, and I will delete it as well.

BTW, I'm Irish (as you might already have guessed!) so I'm constitutionally obliged to wish you a happy St. Patrick’s Day.

Best regards and thanks again for your post.

  • Yeah unfortunately Stack is too big etc to handle what you want to, I think. I'd also not like to hinder students feeling open to admitting they're homework questions because they get too many notices when they mention it. We're not the only ones who go a little above and beyond when "HW" is mentioned (as long as not a bad question). I've seen plenty other users help out. Although, the great "hints" in comments to lead the student to figure/find for themselves is spoiled a bit when someone just posts them the code they need. Not much we can do really. – James Mar 17 '15 at 2:49

You must log in to answer this question.

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