I have been reading a bit of the meta discussion going on about the question quality here at the Physics StackExchange (SE). Especially in the wake of the demise of the TheoreticalPhysics SE, there is a bit of tension between people who want Physics SE to be primarily a place of high-quality/high-level questions, and those who want Physics SE to be much broader. I have sympathies for both camps. Here are some relevant meta discussion:
- Capturing theoretical-physics crowd with Research Level Tag
- Level of questions
- How do I ask homework questions on Physics Stack Exchange?
- What can be done about the (current) ongoing flood of homework and very basic questions overwhelming our site?
- Differentiate research-level questions?
- Where should research-level questions go? Theoretical Physics SE or Physics Research SE?
- Level of questions: what do we do now ?
In light of the continued usefulness of the homework tag, I believe Physics SE should consider emphasizing the research-level tag to help professional physicists extract high-level questions from the noise. I went to the active list and selected the first 7 question that I judged to be research-level, or nearly so:
- Reading list in topological QFT
- Second baryon octet
- Forward-scattering for a single impurity in an infinite system
- Relation of Higgs couplings to masses of fundamental particles
- About the microscopic form of magnetocrystalline anisotropy
- What is crystal field anisotropy or effect ? It forces the magnetic moment to point in particular local direction..
- Does a complete theory of quantum gravity require anthropic post-selection?
I have no idea what would be an appropriate way to emphasize this tag, but I don't think its unreasonable to consider changing the main banner of the website (currently "Here's how it works..."). Other ideas: give it a place in the FAQ, or make it "pinned" to the list of frequent tags.
The current not-logged-in front page of Physics SE has these questions:
- Proof that the Earth rotates?
- Surviving under water in air bubble
- Is a proton collision (collisions like in the LHC) visible to the human eye?
- How can Magnets be used to pick up pieces of metal when the force from a magnetic field does no work?
- Is play-dough liquid or solid?
- Does a car consume more fuel when it's raining?
- What is the Earth truly rotating about/revolving around?
- Why myopic people see this picture differently?
- Why do whips hurt so much?
- Is my boss wrong about our mechanical advantage from our pulley system?
- Can the photoelectric effect be explained without photons?
- Can one black hole suck in another black hole?
- A Musical Pathway
- Noether charge of local symmetries
Note that most of the questions are definitely not homework, but also definitely not of professional interest to researchers. I think most of them are captured by the idea of "popular physics", e.g. "Why is the sky blue?", "Can anything escape a black-hole?", etc. Some of these are great questions, and the internet would be a better place if there were definitive answers available on Physics SE for them, but they aren't of interest to researchers.
(There is a soft-question tag, but this refers to "Questions that ask about some aspect of physics research or study which doesn't involve the actual physics." For example, Q: Where can I find the lifetime of obscure particles? A: The PDG's Review of Particle Physics. Incidentally, this tag appears to almost always be misused as a synonym for "vague or crappy question".)
Dilaton has pointed out that we already have a popular-science tag. Currently, it only has 14 questions tagged. I claim we should emphasize this tag just like we should emphasize research-level. Furthermore, it is important that we develop a reasonably objective definition. (@Manishearth worries that this tag might determined more by subjective tone rather than content.) I propose this definition:
Popular-science questions are accessible to a layman with minimal physics training. They would be appropriate in a newspaper article or non-academic magazine. Such question are distinct from research-level questions, which are of professional interest to physicists. They are also distinct from homework questions.
I have suggested this as the tag wiki excerpt, which is the short-definition of this tag and which appears as a tooltip when someone hovers over the tag. (Someone with 20k is needed to approve it.)
I note that popular-science, research-level, and homework are pleasingly separate (that is, very little overlap) but should ideally capture about 15 % of the questions posted on the site, as the bulk of the questions are expected to be standard good conceptual physics questions that need no additional level tag*. It's my hope that this simple categorization would enable researchers to co-exist peacefully with laymen, which may even encourage a bit of public outreach. In Physics SE's current form, I expect that researchers are actively being driven away because of the difficulty in finding research-level questions.
The research-level tag was proposed here. The consensus at the time is that it was a good idea. dmckee's objected, citing arguments against meta tags as advanced by StackExchange founder Jeff Atwood, but most did not find these arguments compelling.
I reiterate why these arguments fail, in this context: First, a research-level/homework/popular-science distinction is not particular subjective. Second, research-level and popular-science tags could not be used "defensively" to enable violation of the normal posting standards. I think there is a danger of the homework tag being used in the manner, but the consensus seems to be that this danger is manageable. Third, none of these tags are likely to be used pejoratively.
Lastly, I emphasize that since the popular-science tag would be mutually exclusive with homework and research-level, it would not really contribute to any sort of "tag explosion" like you might worry about if we started adding other meta tags like
I change the percentage of questions that are rougly expected to need such a level tag at all from 95% to 15% as the bulk of the questions most probably falls into neither of these three categories. This point was for example discussed in the comments directly below this question, as well as in the comments here and here.
This number means, that ideally about 85% of the questions should have no level tag at all, maybe 5% should be shown to researchers (37 users follow currently research-level) who have a very specialized interest and knowledge, 5% to students who are looking for good additional example problems to prepare for an exam for example (there are 36 people following homework too) and 5% to people who want to learn about physics in a simplified weay without beeing bogged down by math an physic jargon they are not interested in.