Studying theoretical physics in a creative way

[ohannuks]

I have come across a serious problem in my studies. I like to self-study, but I am wrestling between different study methods.

I have at least three options on how to read material:

• Read through it
• Derive every formula yourself (~6x extra time)
• Read the title and outline, and re-invent the bike (~50x extra time)

My problem is as follows: when I just read through something, it will take little time. When I derive everything in the text, I feel it's somewhat a waste of time but occasionally I do get the 'ah-ha' moment. When I read just the outline and try to come up with the methods myself, it takes a really huge amount of time, but I feel like I'm being creative and often I get to the result in some different way than the author/material does. With the third method, I feel I also understand the underlying assumptions better. I also tend to spot more mistakes. The issue is that this takes a huge amount of time.

Having stated my motivation
I would like to discuss different studying methods. What type of studying do you feel is the best for you? What studying method do you think is best for researchers (when you are not supposed to attend to courses anymore and don't have Q&A in the published papers you read)?

Ps. It's somewhat unsettling that I seem to get the best scores from exams when I am going with method 1 (main reason being that the exams are usually based on the reference material)

 

[Geofleur]

I like to read at least three times, each with a different purpose. The first time, I try to understand what is being said and make up simple examples to convince myself that it all makes sense. The second time over, I try to follow all the steps and supply any missing ones. The third time through, I focus on the big picture, asking why the author chose to present things the way she did. During each of these passes, I make notes in the margins and on blank pages of the book, drawing pictures when feasible to represent the main ideas. It's especially nice when an abstract idea can be summarized in a picture or analogy.

 

[micromass]

This is a good question, but not an easy one. It is absolutely true that the third method will teach you most. But it is way too time intensive to really be a good method. As a researcher, you want to get to the point quickly where you can make your own novel contributions. This means spending as little time as necessary on the stuff that is already known (while of course making sure you do understand it very well).

So what is the best? The best is to find somebody who will guide you. Somebody who will tell you "just read this and make sure you understand" and "this is something you should spend your time on trying to figure it out yourself". This is, in my opinion, the best way to learn.

 

[ohannuks]

Hi,

Thanks a lot for all the replies. I'm sorry I didn't reply earlier.

@micromass
That's actually a very unexpected reply, and I tend to agree with you. I self-study quite a lot, so in some sense I was hoping that the best proposed method would involve working things out by yourself. I don't think there's any wrong answer to the question.

I'd love to discuss what it means to be a good teacher some other time.

@Geofleur
I think this is closest to what I've done, as well. Recently, I've tried first reading the abstract or a title and then sketch (quickly without math) how I would solve the problem, then compare with the author's method.