golfpro20050,
Great question! I almost always advise my students to develop a hypothesis before diving into CR questions. My rationale is:
1. If you don't do this, it is often easy to convince yourself other answers (especially the trap answer) is correct.
2. Even if you predict wrong, developing a hypothesis is a quick way to move your brain from reading to understanding, and that step gives you a better foundation for evaluating the answer choices.
3. The more you hypothesize, the more you begin to notice the patterns in CR. For example, the "assumptions upon which an argument depends" tend to fall into a few distinct categories. The more you can familiarize yourself with these categories and understand your own tendency to see or miss certain types of assumptions, the better you'll be able to predict effectively.
Best of luck!
Brett