I believe students need to start with a basic model of research and that is the quantitative experimental model of doing research. That is, start with the outlining of 5 chaptersa
1. The problem, the question, the objetives,
2. The review of the literature,
3. The methodology (design, subjects, materials, instruments, procedures), the data analysis,
4. Results, and finally
5. The conclusion
They will grasp the resarch process easier when they move to different paradigms or designs
Personally I am a believer on constructivists' approach to learning and I take it to research processes. I start by inquiring on what they think is a problem they have found that they think they will like to solve. I begin to explore this no matter how crude they may be understanding clearly that research has a lot to do with individual interests. This is usually a starting point for me and we proceed to see how their suggestions/answers could be scientifically approach. I also make sure that they are involved in this by suggesting some readings. I hope this will be useful.
Dear Leticia Nayeli Ramírez-Ramírez ,I think literature review part is the most essential part to initiate the research project. And it's better to develop a time frame for the project including the literature review, data collecting, thesis writing etc basically. Checking the progress and results of the research by monthly or bi-monthly, is essential to increase the quality of the project.
The supervisor has a responsibility to support and guide students to initiate and continue the research (especially for undergraduate students). It is very essential to be available to discuss the problems which occur during the research period. I believe that some times the strength of the relationship between supervisor and the student result a good outcome and it may helps to build up a better scientific community.