Abstracts are followed by a presentation (poster or oral, also called 'platform'). Poster presentations are best if your project is in a relatively early stage, and you want feedback from people who know more than you about some aspect of the problem. Platforms are great for convincing other people of your 'spin' on a problem. Both give you professional exposure.
You usually write an abstract when you've got enough data to be pretty sure of your conclusions, but not enough to write the full paper. It's important to be far enough along to publish the paper shortly after your presentation in any of the more crowded fields (to avoid getting scooped). The presentation gives you a chance to hear feedback from lots of people, and figure out what issues you need to work out before or as you're writing the manuscript.
As far as the path, it depends heavely on the type of research and the individual lab. Some labs spend a whole lot of time thinking about hypotheses and the perfect experiment before they do anything. Other labs generate tons of data, revise their experiment when the results are ambiguous, etc. Either way, you then analyze the data, write abstracts and/or papers and grants. In basic science, you generally need to have already done the experiments to get NIH grants, and have preliminary data. If you're working with animals, you'll need to get the appropriate approvals from your institution.
The best thing to do if you want to learn this stuff is to go do basic research.
Anka
