I've been thinking about my progress through the years over the past semester, mainly because my style of working has changed drastically since I came to uni. I learned how to make the work easier and how to be more efficient. I thought I might post some pointers here - who knows, maybe a fresher out there reads this and learns from it.
I rarely delete anything from my computer and rarely throw things out - which means that I have copies of most papers I've ever read and a lot of notes. I decided to try and organise it all (in progress, it takes a while to organise over 6 years of stuff spread across several computers and two countries!).
I decided that instead of having dozens of random folders with papers I should create one huge library and keep adding papers to that in the future. I've been slowly adding all those papers and references from notes into a referencing program. I used Bibdesk/Mendeley combo for a while and had a little set back a couple of months ago due to my library file getting overwritten... I had to start from scratch and now I'm using only Mendeley, since the library can be also stored online, which makes it a bit safer. So a note of caution here - it's probably better to pick one referencing program, not mix them together!
Having my references and papers in one place makes things easier. First of all, I don't have to search my entire computer, but can go straight into my library, make a key word search and see all relevant papers - papers that I have already read, many of which are highlighted/with notes. No more "I read it somewhere, but can't remember where".
I found that having a library like that helps me to make links between modules and to integrate the reading I've done for pleasure/during holidays into my university work. Which is a great thing when you are short on time, but want to support a point you are making with a scientific reference - for instance if I needed a general reference about habitat fragmentation I just search my library and quickly recap a couple of papers I'm already familiar with, rather than go and pick a random paper and attempt to read it back to back. It really does help hugely when you are working close to a deadline and need to ensure that all facts in your essay are appropriately referenced. Why not use the directed reading you did in your first year? Why not use those books and papers read in College? Why not use papers found on Twitter and read over breakfast? It speeds the process and makes it easier to get coursework done, but I also love seeing how different bits and pieces come together - I think this is what making progress at university is all about: it is learning to see the connections, to see the bigger picture. It's getting away from doing the bare minimum of work, of having a couple of papers per module - it's bringing it all together and applying knowledge, thinking like a scientist. Soak up all the knowledge you can get.
The above is more about keeping tract of things and organisation, but what about writing itself?
I cannot stress enough - if you make notes on papers make sure that you write down where are those notes coming from for further reference (section or page numbers help too) and write the notes in your own words. It's really important. I find it easier to write notes as I go, on each paper that I read. Then later on I can just pick and choose from my notes, copy the relevant bits out to a new file and edit that. I don't have to re-read the papers, but if something is unclear I can quickly find the sections I should have a look at to clear any confusion. Writing the notes in your own words ensures that you don't commit plagiarism. It's easy to forget what's re-written and what was copied out of the paper, so if you ever copy and paste make sure you have a way of explicitly marking those passages and re-writing them later.
I was also persuaded to give LaTeX a try. I did and I really like it - it's a typesetting system, check it out. It produces beautifully and professionally looking documents, takes care of all the formatting details for you, can be even used to make figures and tables - and I especially like the tables, you can get really nice scientific tables. Some of the scientific journals out there even give you templates, you can download them and then just drop your text in and voila, you have a paper that looks exactly the way it should for that particular journal, references included. LaTeX can also handle references from Mendeley, so I just link the two. I know some people who still do their references for coursework by hand at the end of each essay... This is a silly waste of time to say the least. There is more to LaTeX that the "technical" side too: I find that since I don't have to worry about the formatting at all I can focus on the content and content alone. No more playing with subtitle font for half hour.
I think it's also good to get into the habit of doing all of the above as early in your scientific career as you can - I cannot imagine writing my Masters/PhD thesis in Word! LaTeX deals very well with big documents and a paper library and a good referencing system will be crucial then too I imagine. It's good to stretch and to try out different programs and style of work. Above works for me, maybe it will for you too, but even if it doesn't you should be able to pin point why it doesn't. Once you know that you can move on and try out something more appropriate. Get in the habit of paying attention to detail, but thinking big, of doing more, of reading broadly, even if the papers seem unconnected. It will make you better at what you do, more efficient. It will teach you how to read quickly, but also how to get the most out of each paper, what to look for. With time you will work out your own system of highlighting and note taking, build a reference library and find an efficient way to compile your work. It will make things easier and more enjoyable. Being able to enjoy your work is a truly great thing.