A few years ago, I was asked for some advice from a friend on how best to conduct research. She was in the midst of working on a Masters in Trumpet Performance, and had never had to do any large-scale research before. In response, I wrote the following list of suggestions, reflecting on how I conducted a literature review. Given that I was only a year into my PhD at the time, I'm somewhat proud as to how sensible this list still appears to me. I would be interested to hear if this resonates with anyone else who has conducted this kind of research. Is there anything missing?
Take notes. This might be obvious, but it is vitally important. It could be as straightforward as a skeleton outline of what topics are talked about where. By all means, include page numbers! This will help if you need to find a certain quote, or reread something that didn't seem important when you first ran across it.
Keep a record of what you need to read. It could be as simple as a to-do list, or as in-depth as makes sense without being paralytic. I currently have three lists going, each on a different scale. One is a bibliography of books and articles that I need to read, taken from citations found in other sources. If something looks vaguely interesting or relevant, it gets added to the list. The second is slightly larger in scale, and is decidedly short-term. It deals with projects that are small and can be focused on over the next few days to a week. The third looks at the big picture, dealing with chapters, upcoming conferences, and so on.
Keep a record of what you've read. My system includes date read, location (library, file, online, etc.), full bibliographic citation, and relative importance of each source. There are fewer frustrations than discovering half-way through an article that you've read it before. Likewise, these records make life easier when reevaluating your progress. Further down the road, when you are putting the finishing touches on your thesis, you will thank yourself for having all of the bibliographic information you need already organised on your computer.
Make your own system. Theoretically, any organisational system will be helpful, but it's even more valuable to you if it's one that you've thought out and rationalised yourself. Even though it will take a little time up front to do, and will require tweaking as time goes by, it will prove to be the backbone of your research.
Keep an open mind. In the beginning, you very well might not have a very strong sense of direction – a perfectly natural feeling. As you read, keep note of things that seem odd, unique, interesting… essentially, anything that jumps out at you, regardless of if it seems like a worthwhile avenue of investigation. Balance out your reading with critical analysis, reflecting on what you've read, and compare that to your experiences as a performer, listener, and teacher. It is from those reflections that the really interesting conversation topics will grow.
Talk to people. Whether they're in your field or not, it's always good to bounce your ideas off of someone. They're brilliant for reevaluating your views, and force you to get experience overcoming objections at an early stage. Likewise, you need to be able to explain what you're doing to someone who knows extremely little about music. It takes a lot of mental flexibility to be able to do so succinctly, and will force you to distill all of your ideas into the most simple statements. These synopses, however, will most likely prove to contain the underlying ideas of your work, and are useful in checking your current smaller projects against.
Be critical of your own work. No one can write the perfect thesis from start to finish. There are bound to be sections that you slave over for days, then realise are irrelevant to your overall topic a month later. Don't get rid of them completely, but don't hang onto them simply because you put a lot of time and effort into them. Have several people read portions of your writing and critique it, and learn where your strengths and weaknesses are. From my experience, for every 1000 words I write, usually what I'm trying to get across can be said in 700. Distilling your ideas to their essence and being extremely straightforward in your presentation of them is a valuable skill to have.