What are your favorite study methods?

Getting better at chess is a challenge, just like the game itself.  Some people are unwilling to study at all, whereas others believe that simply playing more and more games will make them better players.  The rest of us try to improve by studying master games, doing tactical exercises, and so forth.  At a basic level, there is not much to say about how to study chess.  Just buy a book on tactics/strategy/endgame/whatever and go to work.  

The more I study chess, the more I think this attitude is wrong.  There are in fact many different ways to study, some more effective than others.

Do you have a special method that is particularly effective?

Let me share a few ideas of mine to give you an idea what I'm talking about.  The problem with studying tactical exercises is that we end up becoming adept at offense but casual or lazy about defense.  One day, as I was working on Richard Palliser's "Complete Chess Workout," I thought of a way to enhance the process.  Once I think I have figured out the offensive manoeuver, I ask myself whether the other player (ie, black if it is "white to move") could partially or fully eliminate the threat if it were his move instead.  Sometimes it is not possible, which is an interesting feature of a position.  But often it is possible, and the challenge is to come up with the best defensive plan.

Let's talk about openings.  Have you been as stupid as I have in buying up opening books, only to learn that (a) it's not a good opening for you in the first place, or (b) it gives you nothing in the way of strategy?  I recently decided to take up the French, but I didn't want to buy Watson's Play the French, which at any rate is (per Silman) intended for 1800+ players.  Instead, I went to Chessgames.com (great site), which charges a small fee for all its benefits, and compiled my own little set of games in the French.  I knew that Gorevich, Bareev, Botvinnik, and Korchnoi are/were notable practitioners.  I also easily found the ECO codes for all main branches of the French.  Using the Game Explorer (maybe this can be done at chess.com just as easily), I looked up all the main variations I wanted to learn, and refined the search to games in which Black wins.  After some copying and pasting, I printed out the score for about 20 "canonical" games in the French, covering the most important variations.  My plan is to play through these games quickly the first time, and then more analytically a second time.  That should give me enough familiarity with the opening so that I can start playing them OTB.  Or so I hope.  More importantly, I will discover quickly what positions feel right for me.  

Endgame study is the hardest, in my opinion.  I don't have any genuine tricks yet, but I like Larry Evans's book "The Endgame Quiz" (something to that effect), because it teaches me the ideas by getting me to apply them.  One other thing I have done is practice simple mates against the computer.  I particularly enjoy mating with two bishops, though I don't really know how useful that is.

What are the best study methods and tricks you've come up with?

  One of the things I started to do once I got the system of notation down is to start playing over completed games mentally without the use of a board.  Seeing through as far as I can go gives me an idea of the level of depth I really have.  Usually I need help after about 10-15 moves (about 5 after an openning) so I know I can think about 4-5 moves ahead.  I've definately gotten better since doing that. 

If I were reduced to a single exercise, it'd be playing over master games slowly while covering up the next move for my side (pick a side for the game).

This is a good example of a nice study method.  I suppose you could do it with a blank chess board, or with pieces on their home squares.  It's important to realize that the ability to see several moves (or "ply") ahead without error is a totally different skill than the ability to select strong moves.  Move selection has nothing to do with it.  So if you are good at sizing up a position, but not good at calculating, this is the kind of exercise you would want to do.