Physics (specifically, physics education of late, though originally relativity), and yes I do.
It’s not unlike solving any problem using the tools that are relevant. You are at point A and you want to be at point B. You have some idea that there’s a path from A to B, and navigating that path might involve some math or calculation or computation. Or you might not be sure there’s a viable path, and you sort of take two steps forward, one step back as you push into the undergrowth looking for a way through.
Vague enough for you?
I think of physics research as fractal-like. While the primary goals are to understand the deepest unknowns of nature, this is the large-scale stuff of the fractal, where real progress happens over years or decades. On smaller scales, months or years, one has to understand things that might not be fundamental to nature but are sub-problems all their own. This might be proving that one calculational technique is equivalent to another one (this sort of thing shows up at all scales, really) or developing a new experimental technique or device (or using one). At still smaller scales, days or months, one might be writing code to deal with a messy bit of math or trying to reduce the noise in some piece of apparatus.
The name of the game is the same at all scales: use deductive reasoning to figure out what the heck is going on. But the actual “thing” you do varies wildly. On the smallest scales, I might need to minimize some mathematical expression with respect to a quantity of interest, so I just set up the math and grind through the first-year calculus. On the largest scales, I might be trying to argue for a new experiment that can answer a question of interest, and I need to demonstrate what the sensitivity is to the physics at hand. This sort of calculation will likely involve breaking the problem down into smaller, plug-and-play pieces (e.g., one that deals with the physics being tested, one that deals with modeling the experimental apparatus, one that deals with how the data will be interpreted, and so on). Depending on the level of precision needed, this might be a week of manipulations on paper or blackboards or it might be two years of writing an extensive suite of software.
This fractal nature of the scientific endeavor is also reflected in the career hierarchy. Though with plenty of exceptions, graduate students tend to battle smaller-scale problems in support of the larger-, longer-scale problems that their advisors are working towards. Postdocs tend to live in the mid-scale with frequent visit into neighboring scales.
Ok, I might have gotten a little away from your question with this belabored fractal metaphor…