First of all, a graduate school is a learning process so finding what is perfect from you from the start is a valiant endeavour but not necessarily sure fire.
You should start to use a data base system, some (mentioned in comments) that come to mind are Mendeley (free), EndNote (commercial), RefBase (free) to mention a few. If you consider going into LateX (which many of us swear by) then I can recommend JabRef (free; BibTeX format). There is also a Wikipedia comparison page that can be of assistance.
To sort your references concerns managing some form of structure where you group or "mind map your articles. In the old days you simply kept them in piles. With a data base software you can start providing key words in the data base and also keep some short notes for each paper which makes everything searchable and ready for sorting. I think the process or sorting articles is one that changes with different tasks and also persons so you are best of taking some advice from others and looking into it. Hopefully you also have peers around who can provide their insights. In the end you will develop your own set of tools that suit your needs and to keep trying different ones at an early stage, and discarding many of them, is far better than trying to do it later when the mass of information is much larger. Another option is to simply start using something and sticking with it regardless of weaknesses.
The bottom line is that with experience you will rely less and less on note-taking and be more efficient at seeing structure in what you read and so keeping a reference database is the main tool you will use. Even if this may not sound very constructive, I also say that the time you spend now on testing different solutions will pay back later, putting it off is only pushing problems forward.
A final personal note. If you want a free, platform independent, and completely versatile way to author documents you should look into LaTeX-writing (for example through TeX.stackexchange). I recommend it to everyone unless you are in a complete Word-environment, being alone with a different system can be hard. You should nevertheless look into it.
One sign of a paper that is not sufficiently developed is that it is organized by source rather than by the author's main ideas:
- First Source
- Second Source
- Third Source
- and so on
A well-developed research paper is organized point by point.
- It is focused on your thesis.
- It uses parts of the sources to support parts of the thesis.
- It uses multiple sources in a single section, because it is drawing ideas and information from various places to support an original idea.
- The same sources will be cited repeatedly in different sections, because different facts or ideas from those sources are relevant to different points you want to make.
- Use a hook to get the reader's interest
- State the question, with necessary background info
- State your answer to the question (your thesis, the point you will try to prove)
- Here you might use part of Source #1 to provide statistics
- Part of Source #2 to provide an interpretation of those statistics
- Part of Source #3 to provide some other facts you need for this major point
- You'll also add your own perspective, tying all these parts together into the single point you want to prove
- Here you might use a different part of Source #2 to provide some other statistics
- Part of Source #4 to provide some other facts
- Part of Source #5 to provide an interpretation of those facts
- Part of Source #1 to provide a different interpretation
- Here too you'll tie all these parts together into the single point you want to prove
And so on for all the points you need to prove to prove the thesis. The Research paper focuses on your own thesis, and uses the sources as needed to provide support for the thesis.
A good rule of thumb: Most paragraphs in the Research paper should cite more than one source. If your paper typically cites only one source per paragraph, thats a sign that the paper should be re-organized.
Think of a car engine: when we do critique (as in the Literature Review), we're not driving the car (what you do normally when you read). Instead, we're popping the hood and taking the engine apart to see how it's made and find the broken pieces. Now, with the final research paper, you've got ten or more engines in front of you, and you're pulling them apart, taking pieces from one and pieces from another and putting them together into a new engine, one you build yourself. Don't just present the reader with one engine and then another and then another. Build your own, single engine from the parts of the others.