Is it me ? or something else ...

Hi all

It seems that most of the questions can be solved easily by Googeling them.

It seems that there is a lot of questions that are home assignments.

I think that the purpose of the the forum is to help each other and not solving homework

sorry for the criticism

Is it me ? or something else ...

Answer 1

[Cedric]

I think that "solved easily using Google" is somewhat an arbitrary statement/judgement, in the sense that there are certainly people who can easily solve your own questions even without using Google, and this is not a valid criterion for stating that it is not legitimate for you to submit these questions here (and the same applies to most of us obviously).

As far as I am concerned, one side of this forum is formation, which means that there is no "too simple" question. For people who are just starting with MATLAB/scientific computing/programming, it is often difficult to filter Google results for extracting relevant information; similarly, it is difficult to search the documentation when they have no clue about how to tackle a problem and no function/command name to start with. So I think that a more fair criterion is the amount of effort that people demonstrate doing once we give them keywords or a few explanations.

In that sense, I have a priori nothing against homework problems, especially when a person honestly declares that it is a homework, and shows that he/she took time to experiment and understand the clues/explanations that we provide. I sometimes also see people who do not understand, but who show that they implemented impressive series of experiments and spent a significant time on the material, and to me these people also deserve additional help.

What annoys me tough, is people who come with an anonymous/fake/one-shot email address, and never come back to provide feedback or to say thank you, and above all people who come back once to erase the content of their initial post. Also people who come and ask for full code to solve their project (some even insist for full code when you ask them what they have done so far: "please provide full code with most efficient techniques for face recognition in dense flow of fast moving vehicles (it must work in all weather conditions and with a cheap camera), and for sending a satellite to Pluto"). I don't remember whether emails are verified when accounts are created, but that could be a good thing to do. Also, I think that it should be mandatory to enter a license number to get a tag e.g. "license verified user" (and then we are free not to take the risk to spend too much time with unverified users), because if people knew that Mathworks knows in which institution they are working/studying, it could dissuade them to misbehave on the forum.

[Jan]

There are many active users who are not associated to a license, e.g. the user of a university license like me. And Octave users ask good Matlab related questions also. But in both cases there might be other unique signals to estimate, if the user is really interested in the answers. And I am really interested in the code about the satellite transfer to Pluto. The name of the (mini-)planet must be an input argument as string. Send it to me, asap.

[Cedric]

You have a license number for your university/department; my point is that it could be dissuasive enough already if Mathworks knew the institution, and forum users would have this information reminded at the moment they post something (EDIT: the time and the IP address could be displayed as well).

About the satellite, I am working on a more modest algorithm for sending myself to the nearest coffee machine. I'll send it to you as soon as I implement the landing part (right now I am still stuck orbiting around the darn machine).

[James Tursa]

Pretty much in agreement with all of your points. I don't have a problem answering homework problems, but only after the poster has demonstrated a sincere effort. And even then I mostly correct what is posted and suggest improvements w/o giving complete answer. It does annoy me when posters delete the question, an obvious attempt I think to cover any evidence that they didn't do all the work themselves. For that reason I wouldn't mind a restriction that posters cannot delete or edit their own questions once they have been tagged as "homework".

[Jan]

@James: Posters without editor power cannot delete their question after an answer has been posted. But they should own the possibility to edit the question, because this is the best location for answering questions for clarifications. If important information is hidden in a comment (and out of sight after new comments are added) or in an answer, reading the question does still confuse new readers of the thread.

Displaying the IP address is critical, because this conflicts e.g. with the German laws of privacy protection. It is in many cases not a big deal to find out the personal identity corresponding to an IP and then e.g. a housebreaker can use the chance to control when I'm at work.

Answer 2

[Matt J]

It's not you, but contributors here with appropriate privileges are doing their best to filter doit4me homework posts. The situation is worse over at the Newsgroup site where none of the frequent contributors have any filtering power at all.

Answer 3

[Jan]

I estimate, that less than 10% of the questions can be solved by Google. But this is still a surprising high rate, when we consider, that posting a question takes more time and waiting for the answer lasts much longer.

The homework questions are tedious. Unfortunately too many of them are answered.

The purpose of a public forum is defined by the contributors also. This crowd cannot be controlled easily. We can add the "homework" tag, ask about the own effort, suggest using Google and more frequently the documentation of Matlab.

But I see no spam and off-topic threads here. The tone is friendly and constructive, which is an effect of the power of the editors and admins to delete rude messages.

26'000 answers have been accepted. Even when the homework questions are removed from this number, this is a nice piece of work for 924 days.

So your criticism is welcome and you mention facts. Do you have a suggestion for improvements? What about letting the user confirm by three check boxes, that s/he has asked Google before, read the help texts and does not post a homework question without showing the own effort?

Matt Fig has written: "Answer questions about Matlab - the rest is fluff".

Answer 4

[Image Analyst]

Sure, there's Google. Sure, there's the FAQ. Sure there's the File Exchange. Sure there's the Mathworks Search. And sure there's a fair number of "no attempt yet" homework problems and "I need code for [some complicated thing]. Please send me complete code." questions posted. And lots of questions where people don't know how to format their code, and can't explain their question in clear English (or even in broken English). But that's the nature of the forum and human nature. If I didn't like it, I'd leave. But I know I'm helping a lot of people because they tell me so, and I see snippets from my examples/demos posted in questions, so that's enough to make me stay.

Answer 5

[Evan]

Others have addressed the other topics of this post, but I would like to comment about what I see as "reasonable expectations" for filtering homework questions. It might not be completely on-topic, but I feel it's important.

Oftentimes, it is explicitly clear that a thread is a homework question. Many users either state that they're working on homework or copy and paste their assignment in its original format into their post. It's also often quite clear that the user asking the question has no interest in producing a solution themselves and instead wants something that they can copy and paste into an .m-file for submission. In these situations, I see nothing wrong with letting the user know that they won't receive the type of "help" they're asking for.

That said, it requires little effort to mask that your question is about homework. Because we're not in any position to know the user's intent for the code on which they're working without them explicitly stating it, we are to some degree going to be giving out what some instructors find to be an inappropriate level of assistance. I think this is inevitable, and don't think we should be involved in the guessing game that is trying to figure out who is cheating on an assignment and who isn't. While it's not an ideal situation, classes should be set up with enough exams or large projects that you can't get by just by copying some simple scripts from an online forum, anyway.

Unless we have clear evidence that a question is about homework, we should assume it isn't. If we try to play detective, we'll end up punishing users for the crime of posing their problems as simplified scenarios that end up inadvertently looking like something out of a textbook or assignment page. I look back at some of the early questions that I posed here, and the examples I gave that demonstrated my problem were so simple, and the problems themselves so basic, that were I looking on them now I would be tempted to think past me was asking about homework.

This means that some users will able to get by with having an unfair level of help just by saying "this is for a job" or "I'm just trying to teach myself with this example." It's unfortunate, and we could perhaps mitigate it by providing generalized answers + an explanation of each line of code, but in the past I've seen people be outright refused help because everyone assumed they were working on homework, even when there was no convincing evidence to support that.

[Matt J]

It's also often quite clear that the user asking the question has no interest in producing a solution themselves and instead wants something that they can copy and paste into an .m-file for submission.

This really is the better criterion for offering or withholding help, regardless of whether it's for homework or not. It is also pretty easy to identify.

[Evan]

Yes, I agree. I probably should have made more clear that I agree with Cedric's philosophy on problems that are clearly homework. In short, whether it's homework or not shouldn't greatly affect how we handle it, and we shouldn't be in the business of interrogating posters about why they're asking their question or making assumptions about their intentions.

[Jan]

Many time questions are tagged with "homework", when they look suspicious. As soon as the OP claims, that it is not a homework, this tag should be removed. But this does not happen usually. The questions, which are posted with as less own effort as possible are not maintained by the authors and questions for clarifications are left unanswered frequently. Therefore I think, that asking the OP if this is homework, is a trustworthy method in 98% of the cases.

But the homework questions are not an important problem for me. While a (set or imagined) DoIt4Me tag is a secure reason to ignore the question, homework means only, that I do not post a full solution, but answering a specific point is still a "fair use" of the forum. E.g. mentioning the required command and point the user to read the help text is no cheating.

I've sometimes posted the comment: "Please do not solve homework questions." Then the corresponding answer has been deleted by the author in 9 of 10 cases. This was encouraging.

[Walter Roberson]

I make no special effort to remove inappropriate "homework" tags. If I happen to read the posting at a later time and see that it is tagged but is not homework, then Yes, I will remove the tag, but I do not go back and review postings with the tag to see whether the tag is still warranted. I have not seen many cases in which a "homework" tag turned out to be incorrect.

More common, in my experience, is that a "doit4me" tag is appropriate based upon the initial posting (and possibly a couple of rounds of back-and-forth afterwards), but in time the poster demonstrates that they are working on the problem and not simply expecting someone to solve it for them. Removing the "doit4me" tag would likely be appropriate in those cases. I will do that if it happens to come to mind while I am looking at a post, but I do not do specific reviews for this purpose.

[Walter Roberson]

For homework issues, I have many times gone to fair lengths to not post solutions, and to instead lead the poster through steps in understanding until the poster understands how to ask the right question, and understands the steps towards solving the question. The number of times I have clenched my fists and gritted my teeth to stop myself from posting, "You can solve this entire question with this one short line of code: ...". Give a man a fish-stick and he probably won't even notice what he is eating.

In cases where someone is working on something and the step they are missing is clearly a side issue rather than a lack of understanding or work, then, Sure, I don't mind giving the code outright. Even if it turns out to be the same code that was the desired "learning experience" for the previous person's homework.

[Jan]

@Walter: Even for non-homework questions you provide strategies, and not only bare code. Therefore other readers like me can profit from your answers also and your answers support the total quality of the forum. Therefore your work, and of course the work of others also, increase my motivation to participate. Sorry for the flattering, I don't think you need some candies, but perhaps other readers of this message are encouraged to learn from how you post solutions.

Homework questions are no problem, but the quality of their answers are.