This pitfall suggests that we need to open up spaces for our students to experiment with material as a means of learning and should reject the banking theory of knowledge that dominated education in the modern era. I’m fairly sure this idea works well in more creative endeavors, but statistics at the undergraduate level does not provide much room for stumbling on to truths (provided one wishes to finish in 10 weeks). To this end, I feel that students may require a set of boundaries to their play areas that allows them to work with ideas and then move forward.
In my course, students are working on an original piece of research, so they have selected a pair of concepts in a relationship and then must figure out ways to measure these concepts, describe their data, and ultimately test the relationship. While there is a most right answer to measuring each of the concepts with the data provided, by letting students stumble around the right answers while keeping them in touch with their classmates looking at similar ideas, students should have the experience of being allowed to freely play with the information while not actually being subject to the multitude of rabbit trails unsupervised researchers may follow.
This can be further facilitated by setting up discussion groups based around specific concepts or statistical methods to allow students to pool their knowledge; however doing so does allow for the pooling of ignorance as well. This reality will require a deft hand in balancing the need to let students explore on their own and make mistakes while keeping them sufficiently on track to finish the course in time.