http://pchsfysicstchr.edublogs.org/2012/03/11/the-problem-with-problem-based-learning/
I like the idea of problem based teaching, since it gives the students more control over what they're learning and why they're learning it. That being said, it's up to the students to come up with a research question so they can gather the necessary sources and materials. The issue here is for teachers to help students identify a research question that is both challenging and feasible, and given the amount of trouble people had in my PhD days, this will be very difficult.
I do agree with the author of this post that teachers must continuously learn in order to properly guide their students; my advisor was constantly reading current literature in order to formulate new questions for upcoming grant proposals, and likewise, I had to read literature in order to find sufficient background information for my project. However, that was a doctoral program, and much of the pressure to design the experiments and search for relevant journal articles was on me. When I first started out, my advisor provided me with a few articles to kick start my reading and build my background knowledge on the lab's research interests. There was quite a bit of hand-holding as I somehow managed to stumble through my first year, and eventually found a good routine to reading paper efficiently, taking notes, and coming up with possible research questions. After I passed prelims, things started shifting and it became more on me to come up with ideas, while my advisor's role became more mentor-like rather than lecturer-like.
At the high school level, I would imagine it to be more like the first year of grad school, with less restrictions on what counted as reliable resources. While I was stuck using PubMed and books lying around the lab, there's nothing stopping a high school student from searching the textbook first for some background information. The teacher then becomes more like a consultant, asking the student about how they propose to continue their research, rather than the source of all knowledge.
The author talks about the difference between "ill-structured" and "well-structured" questions, and while most scientists, not to mention grant agencies, would like to see well-structure questions, I don't see anything wrong with starting with an ill-structured question, breaking it down, and the assigning the well-structured components to different groups. After all, scientists do collaborate...
