11

I've seen a lot of answers where it's just a random idea, with no way of telling whether this is a good tip, an old wives' tale, or something the answerer just made up on the spot.

For example:

Put the onion in the freezer for 3 minutes. It works wonderfully.

https://lifehacks.stackexchange.com/a/235/148

Put them in the microwave for 10 seconds on high prior to the cutting.

https://lifehacks.stackexchange.com/a/259/148

Ants don't like cucumber. Keep cucumber skins on the shelves and it will stop ants coming to food items.

https://lifehacks.stackexchange.com/a/237/148

These answers are extremely short, but a bigger problem is that anyone could just post any old thing and you have no idea, without testing, whether the hack actually works. Urban legends and old wives' tales could be perpetuated, and we would not be a reliable source of information on the web.

Note that I'm not saying that any of the above tips don't work; the point is that because no information has been provided, there is no way for a random reader to tell if they work or not.

7

My personal thoughts are that answers should contain one or more of the following in addition to the actual "answer" part:

  • a citation to a reliable source that says the tip works
  • a link to something (e.g. a video) that shows the tip working
  • an explanation of why it works

This means that if you can say "putting the onions in the freezer works because it denatures the enzymes that react with the onion juices to create the gas that makes you cry", then you don't need to necessarily also provide a source for that (though if you have one, that's great!).

I think that requiring at least one of these in an answer would improve the quality of the site and help draw in the experts.

  • Agreed. While I'd prefer an answer that posted a scientific explanation also link to a reputable source of that information, I would not penalise one that didn't (although I'd choose to lend my vote to an answerer who did), provided the explanation was thorough. – Nick Udell Dec 11 '14 at 11:26
6

Here are my basic guidelines that I follow for writing answers:

Step 1: Explain why they're having the problem, or what causes the problem.

Even if it's a bit redundant to you, a quick summary could be useful to someone who is reading the question and not having the same problem. It helps to create a point of reference so that everybody is on the same page, and allows you to lead the reader into agreeing that your answer is the correct solution.

At the very least, if you're having trouble making your answer more than one or two sentences long, a summary will add fluff to your answer, saving it from being criticized for being too short (No promises on not being criticised for being fluff though).

Step 2: Explain how to fix the problem.

This is where your answer lies. Depending on the question and your answer, this could be the shortest or longest section. Here you simply state what your solution(s) to the problem is.

Step 3: Explain why the solution(s) is/are the answer to the question.

The most important part of the answer, this section is where you convince the readers that your answer is the correct one. If you don't include this section, all you're doing is making a statement. Here is where you can explain how you experienced the same problem until you used your suggested solution, or provide references to show that your solution is the correct one.

If you're having trouble writing this section. Try writing it as if your were responding to the question: "Why should I believe you?".

  • 1
    +1, but I think that #1 can be left out at times if the question provides a good description of the cause – Zach Saucier Dec 11 '14 at 16:27
  • I believe #1 is a good thing as it does enforce that the question is understood properly, which aids in not answering an entirely different question – holroy May 31 '15 at 19:57
4

We should upvote high-quality, well-put answers while downvoting / encouraging users to improve low quality answers as to give out a signal that we treasure better answers.

This sounds like guess work, the answer could be improved with references. – James Jenkins 2014-12-10 12:51:46Z

  • The problem with your answer is that it relies on each individual's idea of what a good answer is and isn't. It kind of goes without saying that we upvote good and downvote bad. But the first example I gave in my question has 6 upvotes, and as for the third, when I suggested improving it, I got a "?????? .....who said citations were needed here?" reply from @Shokhet. So we need to come to a community consensus on this, including defining what exactly is a high-quality answer and what isn't. – starsplusplus Dec 11 '14 at 11:34
1

Not downvoting and not leaving negative comments to discourage non-hacky answers would be a good way to improve answer quality.

In my view, conventional solutions using existing, well-established tools and methods are implicitly verified, and so discouraging them discourages known working solutions, attracting a higher proportion of unverified answers to begin with. E.g. with the ant example from the OP, conventional solution is identify the type of ant, remove food sources, use appropriate baits/pesticides, call exterminator for larger infestations. That works. Discouraging this encourages unverifiable, or at least less effective, things like "ants don't like cucumbers / cinnamon / peppermint oil / etc.".

Often times the highest quality answer is an easy conventional solution that the OP was not aware of for whatever reason, while the hacks are frequently misguided, unsafe, inconvenient, or otherwise less than ideal.

One recent example is How can I remove ants from my laptop keyboard?, for which the OP actually found the best solution in chat (just use ant baits) and was then hesitant to self answer because it wasn't hacky enough:

OP: I ended up following ton.yeung's advice in a chat room and got rid of them just fine, now I have no idea which answer to accept since I didn't try any.

Person: cw self answer quoting the chat advice

OP: It's not really a hack, just bought a box of ant thingy and left it on the table.

This isn't an issue with the OP, but rather a bigger picture issue resulting from the actual best answers being discouraged because they aren't hacks. The exchange quoted above is a big time red flag. Many, if not most, of the questions here are answerable with existing tools that are designed specifically for the task being asked about. To not recommend such tools (tools that were generally created after much struggle by others in the past with the same problem) is lower quality advice, by nature; such answers do not build on the experiences of the past.

I acknowledge that the question of whether or not "conventional" answers are acceptable is controversial. I also acknowledge that even the definition of "conventional" is relative and up in the air. Such discussions can be readily found on meta and I understand them. Regardless, the fact is that discouraging "conventional" answers places a constraint on many (not all) questions that leads to low quality hack-style answers.

It must be accepted that just because it is a hack doesn't mean it's a high quality solution to a problem. That isn't necessarily bad, per se, it just means that if conventional answers are to continue to be discouraged, then lower quality solutions must be accepted as a necessary consequence. Which is fine, if that's the goal of the site.

  • Did you read the body of my question, or just the title? You raise some good points, but it doesn't really answer the question of how to stop unverified answers. – starsplusplus May 31 '15 at 17:42
  • @starsplusplus I did read the body. I guess I should have made my connection clearer. In my view, conventional solutions using existing, well-established tools and methods are implicitly verified, and so discouraging them discourages known working solutions, attracting a higher proportion of unverified answers to begin with. – Captain Obvious May 31 '15 at 18:06
  • (@starsplusplus E.g. with the ant example, conventional solution is identify the type of ant, remove food sources, use appropriate baits/pesticides, call exterminator for larger infestations. That works. Discouraging this encourages unverifiable, or at least less effective, things like "ants don't like cucumbers / cinnamon / peppermint oil / etc.".) Better enforcement of "does not need a lifehack" close reason ties into this as well, but appropriate use of that reason is too controversial right now to be effective. – Captain Obvious May 31 '15 at 18:10
  • Ah, thanks for clarifying. I agree that existing solutions are implicitly verified since they are designed for exactly the purpose given. I wouldn't say this by itself is a solution for verifying the non-traditional answers, but I guess I see your argument that discouraging the traditional answers could be contributing to the problem. – starsplusplus May 31 '15 at 18:31
0

Thorough instructions of how to execute one or more methods, preferably with images, videos, and references to supplement text instructions.

Answers should also include why something works.

  • 1
    How does this answer differ from what starsplusplus suggests? – Zach Saucier Dec 11 '14 at 14:28
  • 1
    @ZachSaucier "Thorough instructions of how to execute one or more methods"; links are to supplement text instructions in the answer itself. – Mooseman Dec 11 '14 at 14:31
  • Good answer. You might mention links in your answer itself – Zach Saucier Dec 11 '14 at 14:32
  • @ZachSaucier Done. – Mooseman Dec 11 '14 at 14:33
  • This is unclear. Are you suggesting every answer needs to have this? Are you suggesting it as a fourth "one or more" to the suggestions in my answer? Something else? What's the actual suggestion here? – starsplusplus Dec 11 '14 at 15:51

You must log in to answer this question.

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