Do you do all the problems in a textbook?

[Happiness]

Take for example Griffith's Introduction to Electrodynamics. In every chapter, there are roughly 35 questions that are embedded as part of the text and another 20 questions after the chapter.

What is a good guide? Do you do all the problems? If not, which problems would you choose to do and why?

 

[micromass]

Depends on the book and the quality of the problem. For example, take an introductory calculus book like Stewart. It would be insane to solve all the problems. After a while you feel you mastered it, and you moved on. Then there are more advanced books (which usually have a smaller amount of problems), usually all those problems will be worth considering. And then there are the really advanced books where each problem is a small research project.

You should certainly be able to solve all the problems in a text, but that is different from actually solving them all.

 

[PWiz]

When I'm learning new concepts in a textbook, I like to solve a few easy questions at the end of a section after a complete read through to get going, and then skip to the last few difficult ones to finish off the section with the satisfaction that I understood things well without investing too much of my time in problem solving.

When I'm learning material before an actual assessment though, I like to be as thorough as time permits me to be and do almost all the questions that I come across, except for the first 10% of any exercise, which is usually composed of trivial problems.

This is a bit subjective of course - the most difficult problems in any text are usually a notch above the most difficult questions you're likely to encounter on a test, and doing all of them is not really necessary. But if you want to go for perfection, hard problems are the way to go.

 

[clope023]

Take for example Griffith's Introduction to Electrodynamics. In every chapter, there are roughly 35 questions that are embedded as part of the text and another 20 questions after the chapter.

What is a good guide? Do you do all the problems? If not, which problems would you choose to do and why?

Well the ones assigned by your professor are usually a good starting point, but there's no reason to do EVERY problem in EVERY chapter. Extra problems beyond your assigned homework ought to be focused on your own conceptual difficulties with the material; you only have so much time to study between tests, so you should use your time efficiently.

 

[Incand]

I think it can be good to do all in some courses depending on how much you learn from doing problems as opposed to doing the text and how interested you are in the subject.
For example I did every single problems in our first course on classical mechanics (maybe about 200-300 questions) because that was how I really felt i learned. Understanding the basic principles is rather easy but knowing when choosing one over the other to get simpler calculations is all about practice.
For introductory thermodynamics I did slightly less but still a lot more than was recommended while In waves & optics I only do the recommended ones because most of the topics are somewhat familiar from high school and optics doesn't really interest me that much.

But it's important to note that this is for your own learning, if you just want to get a good grade doing a few exercises and then practicing old exams and assignments (assuming they're somewhat similar each year) is probably a lot more effective.
Another reason for me personally to do more than the recommended ones i that I'm at a rather low ranked university (I could've gotten in anywhere but choose something small and close to home) and to get an A I would only need to put in 10-15h a week but I want to learn the material as well as if i was at a top school with higher pace.

 

[neosoul]

I try to do as many problems as I can. Then, I get a more advanced textbook or go to my teacher for more challenging problems.

 

[Vanadium 50]

You should be able to do every problem. This doesn't necessarily mean you have to do every problem.

 

[Dishsoap]

Take for example Griffith's Introduction to Electrodynamics. In every chapter, there are roughly 35 questions that are embedded as part of the text and another 20 questions after the chapter.

What is a good guide? Do you do all the problems? If not, which problems would you choose to do and why?

For Griffith's E&M, the more difficult questions have stars (or circles, depending on the book version). I would say first when doing a chapter look through and find the non-starred questions, read them, and make sure that you know how to do them. If you aren't 100% confident in the question, try and solve it. Even if you think you can solve it, you might also want to look through the solutions manual to make sure that you aren't missing any nuances in the question. If your course is rigorous such that you need to know how to solve the absurdly difficult questions, move on to starred questions.

Also, for some reason, Griffith's E&M is absurdly varied in the number of questions it presents per chapter. It might be ~35 in one chapter, but in one chapter (I know chapter 12 is this way), there are upwards of 75 questions. So for sure, don't do all of them.