When I was in college:
For physics, more. Granted I took a more advanced courses my first year (the entire course was mostly juniors/seniors). I would say about 10 hours a week once I average out my “all in before exam” days.
For math in my first year including Honors Linear Algebra and Modern Algebra, I would say more. 5 hours a day (I never had proper exposure to proofs so it was the only real way to catch up) for the 1 class seems more reasonable.
For machine learning, I would say about 35~45 hours a week? Think I scored the highest in class in the essay portion of my midterm so probably not a good example especially when considering the school I attended is currently considered top 3 in US News undergrad national ranking.
For philosophy (and many of my humanities courses), I would say 0~2 minutes a week at most? Somehow, I received A’s in some of those courses.
For more difficult humanities courses that required 15 page essays routinely every month or two with quite a lot of reading, I would say 1~3 hours a week at most and a “one nighter” (granted this is a poor way to write an essay but procrastination is a thing) with essays each time?
Basically, for me, it swinged widely. Some courses were full time jobs in themselves and other courses required zero effort in my part.
There were definitely courses I felt I just didn’t have the time (and maturity in thinking) to do well despite the number of hours I put in outside work-study.
But I believe this isn’t the norm. I had sections in which I had one of the highest if not the highest scores in a class of 100~300 students (and mind you, the school I attended was very reputable). Plus, college isn’t high school. The chance of being the top student in your course is rather slim. Don’t go crazy over it. I didn’t see much differences between a student who earned a B~B+ and a student who earned an A. The grading felt arbitrary with curved exams. Also, I felt many times some students who were getting Bs to have a stronger grasp of the subject than the students receiving As. (Not the norm but something I evidenced myself which helped me realize grading in STEM was pretty stupid. It favors test taking skills over utilization of knowledge. A few students who got Cs could have gotten it because of a poor exam score. A student who had inadequate knowledge of the subject could have gotten As because of a lucky final exam score. And then there’s the issue of rampant cheating and all in today’s universities which seriously penalizes integrity (in the grading scale).)
I would say, study enough so that you feel challenged and you are either understanding the material fully or trying your best with the time available to do so.
It is much more satisfying to get a B or C or whatever after studying in a course basically full time over getting an A in a course that one puts zero efforts in.
However, I did notice for “career” purposes, taking easier courses for "A"s were better looking better. But I have no regrets personally. Up to what point would you fake yourself to suck up to society?
College is a place to train your mind. Challenge yourself. Just do your best.
That isn’t considered good GPA by most people even in STEM. UCB is more correct what a “good GPA” is though 3.0 seems just an average GPA, not a ‘good’ one. A good GPA in my perspective outside med and grad school is around 3.4.
If you consider med school, 3.7+. If you consider phd, probably 3.8+ with researches, etc…