It is a common criticism of major research universities that professors don't spend enough of their efforts on teaching. I don't disagree that these are sometimes valid concerns, but I do believe that research and teaching can go hand-in-hand. One argument can be made that the people who are at the forefront of research in their respective fields are best qualified to teach students of those subjects. However, I think a more important notion is that research and teaching are both similar, iterative processes. When I conduct an experiment in my research, I study the outcome and develop recommendations for how to proceed or design new experiments. When I teach a course, I similarly like to look at how the students respond, and I adjust my techniques for the proceeding courses or lessons.
During my graduate studies I had the opportunity to teach individual class sessions for my advisor's semester-long course in Analytical Product Design (APD), which is a project-based course where teams design and build a solution to their problem of choice. Most people use the term "lecture" to refer to teaching a class session, but I find that giving a lecture is a much more restrictive activity than leading a class session. A lecture implies that the instructor speaks while the students listen, whereas a class session, to me, simply refers to a specified amount of time during which the class meets and the students learn. I had the privilege of teaching 2-5 sessions per year for four years at the University of Michigan, covering topics that include survey design, prototyping, microeconomic analysis, business planning, and design optimization.
I came into this with a bit of experience teaching at the middle school, high school, and undergraduate levels, and I had some ideas on how to engage the class during our sessions. My first APD classes were designed to be very interactive: I presented a concept through class discussion, and then the teams would have time to discuss how to implement that concept for their term projects, followed by a full-class discussion so that different teams could learn from one another. Kind of a variation on the think-pair-share technique that leverages the team structure of the class. While this teaching style let me see the student teams in action as they applied the course content to their projects, at times it felt like I was micro-managing the team discussions by telling them what to talk about and when. The next session I led was designed to move more quickly through the course material and leave time at the end for team discussions, but this technique left me feeling as though some of the students weren't engaged while I presented the content. In subsequent sessions I sought to converge on an appropriate balance of these styles, both engaging the students through discussions about their term projects while not over-doing it, leaving time at the end for more general team discussion.
I see each class session I teach as an iteration on all of my past classes and experiences. Even when I am tasked with teaching the exact same content year after year, I always take the time to reflect on my teaching style and try to improve for the latest iteration. This way, I as the instructor often feel like the student, learning new things every time I teach.