In my mind, ideally, there should be a fairly consistent "etiquette", if you will, for choosing tags wherein any two people asking the same question are likely to pick the same tags for that question. With this in mind:
What is the appropriate use of the analysis tag?
This tag's description reads:
Mathematical analysis. Consider a more specific tag instead: (real-analysis), (complex-analysis), (functional-analysis), (fourier-analysis), (measure-theory), (calculus-of-variations), etc. For data analysis, use (data-analysis).
If I were to ask a question about, say, Bolzano-Weierstrass, I'd choose the real-analysis tag. If I were to ask a question about Cauchy's integral formula, I'd choose the complex-analysis tag. Once I do so, it seems superfluous to add the analysis tag on top of it. Should I? If not, are there any scenarios in which one should choose the analysis tag over one of the more specific tags listed in the description above? If the answer is no, then why do we have the analysis tag at all? In practice, I see this tag used quite inconsistently:
- Some people add both the analysis tag as well as a more specific tag.
- Some people add only the more specific tag.
- Some people add only the more general tag.
One reason this is problematic is because some users browse by tag instead of clicking "newest questions". Moreover, it makes it harder for those who use the "favorite tags" feature to find questions that they'd be interested in answering. I'm personally unwilling to add it to my favorites because then I'm sure I'd get a lot of "spam" highlights from harmonic analysis or functional analysis—topics I know nothing about. But if I don't add it, then I'm passing over questions I'd like to answer.
Addendum: In light of John Ma's answer, I propose that we edit the tag description to say "Please use a more specific..." instead of "Consider a more specific...".