Trainees work on scientific problems with collaborating scientists. They learn by doing and are mentored by a senior scientist.
You will need
- An idea for a research project that is mutually acceptable to the supervisor and the trainee
- Funding for the project
Benefits and limitations
- Participants receive deep training; sometimes to the extent of becoming a globally recognised expert and/or innovator in a particular area
- Participants acquire both technical skills and soft skills, including time management, project management and communication skills
- It is time consuming and thus does not scale well (a typical supervision ratio is 1 trainer to 4 trainees)
What do I actually do as an instructor/facilitator?
Preparation / before class
You need to define a project relevant to your group's research interests and of an appropriate length for project-based learning. Typically, 6 months to a year is considered the shortest period of time for such a project to be beneficial both to the traineee and to the group of the supervisor. Often such projects are longer - 2-5 years. You or the potential trainee need to secure appropriate funding and ensure that the facilities, equipment and learning environment are appropriate to achieve the intended goals of the project. You need to identify candidates with the appropriate prerequisite knowledge base and, most importantly, with the potential to self motivate, work independently, and interact productively with other research scientists.
Implementation / during class
Project-based learning can be applied to the supervision of any predoctoral or postdoctoral scientist. There is a strong emphasis on training, not just on scientific output. Your goal is to facilitate the development of a well-rounded research professional, with deep knowledge in the problem domain the skills to see research projects through from design to publication, and the ability to perform successfully in scientific collaborations. Much of this needs to be driven by the learner, with your support as a supervisor.
Sims D, Ponting CP, Heger A (2015) CGAT: a model for immersive personalized training in computational genomics. Briefings in Functional Genomics 14: 1-6 (doi: 10.1093/bfgp/elv021)
Examples and materials
Computational Genomics Analysis and Training (CGAT)
Photo reproduced with permission from Novartis Pharma AG