I began my second-career as a teacher using a "traditional" teaching style. This was the way I learned and I did not have formal training in education. My teaching style has evolved to become more constructivist. My masters courses have provided the favorable research and I have seen it work well in practice. Doing is more powerful than hearing and seeing. Teaching using constructivism is not easy. It requires lots of forethought about the learning goal for each lesson and the many paths students could take to reach it. I have become better at providing the different degrees of scaffolding necessary. This varies by class, grade, and concept. I am also getting better at creating real-world problems to drive discovery. Once in a while, I found myself lecturing and felt guilty. I stumbled upon a white paper that helped sooth my sole.
A Time for Telling examines when lecturing is worthwhile and when it is most successful. Constructivist teaching gives the students substantial control of their learning, whereas, traditional teaching is controlled more by the teacher. In constructivist teaching, the teacher still plays an important role of guiding discovery through scaffolded inquiry. Lectures and speeches are meaningful when they "map into the knowledge of problem situations." Think of a coach that gives a meaningful speech at halftime. A successful speech will be motivational and tie into the team's strategy and game plan. "When telling occurs without readiness, the primary recourse for students is to treat the new information as ends to be memorized rather than tools to help them perceive and think."
A less effective speech will ignore prior knowledge and contrasting cases. No amount of additional "telling" will improve the comprehension. In fact, novices who lack proper background can actually think they understand the concept, when they missed important distinctions. An example would be the different levels of details understood by a novice (lay person) and an expert reading an article about a baseball game. "Contrasting cases help people notice specific features and dimensions that make the cases distinct (Bransford & Schwartz, in press)."
The research had students analyze contrasting cases and then were tested one week later on the hypothetical outcome of an experiment they did not actually conduct. The contrasting cases were compared against more traditional studying techniques (e.g., summarizing, re-reading, etc.). "Generating the distinctions between contrasting cases and then reading a text or hearing a lecture led to more accurate predictions than the control treatment of (a) reading about the distinctions between cases and hearing a lecture, (b) summarizing a relevant text and hearing a lecture, and (c) analyzing the contrasting cases twice without hearing a lecture. Interestingly, the benefits of analyzing the cases extended to concepts that students did not actually discover in their analyses."
So, lecturing can be an effective teaching technique, but only when it extends students' prior knowledge. If students lack prior knowledge on a subject (e.g., they are novices), then the lecture will not make deep connections. Students can develop prior knowledge by analyzing contrasting cases and synthesizing similarities and differences (e.g., discriminants). Teachers can "give" students the prior knowledge by pointing out important features of the cases, but this is less effective than when students discover them on their own. "It is relatively easy to tell a distinction to someone, if that person shares the same set of experiences". This discovery process does not have to be a full-blown activity. Rather, it could just be a brief collaborative activity for a few minutes before a lecture. An important detail of these activities is that they differ from projects. Projects often focus on the final project. Discovery activities to build prior knowledge emphasize the process and often don't have a final project. The discovery process should be guided. For example, before introducing a math units on solving distance/rate/time problems, provide students with several different models that can be used to solve this type of problems.
"Although contrasting cases are effective at scaffolding the development of differentiated knowledge, there is a limit to what we can reasonably expect people to discover." In practice, the discovery process takes more time than lecturing. I like to use scaffolded discovery for concepts that are complex or not intuitive.
 Schwartz, D.L., Bransford, J. D., A Time for Telling, Cognition and Instruction, Lawrence Erlbaum Associates, 1998
 Legerstee, Maria, Infants Sense of People, Cambridge University Press, Cambridge, UK, 2005
 image from link