(1) I piloted a new book, one that was more accessible to the students but still had mathematically rigorous problem sets;
(2) From the first day on, once each week, I spent some time talking about how I was organizing the course and why;
(3) I explicitly stated, each time I gave a problem-solving demonstration, ``This is the part where you take notes and ask questions of me about the material. When we get to the activity, you ask each other first, then come to me.''
After reading the midterm evaluations in Fall 2003, I learned to do the complement of Item 3 above: point out to students when we were working on an activity that was directly connected to the concepts they would read about in their book and engage with in doing homework problems. It is reasonable to conclude that the positive shift in student perceptions of the course between Fall 2002 and Fall 2003 were influenced by my acting on the Midterm Evaluation information.