By: Maristela Petrovic-Dzerdz, EDC Instructional Designer
What Prof. Nimijean presented at the last EDC roundtable is a very interesting example of applying two different instructional approaches and getting a possible “Goldilocks dilemma” where things might end up “just right” using a third, mixed approach.
In the first version of the course, live lectures were supported with text rich PowerPoint slides that served as lecture notes to students. Prof. Nimijean revamped the second iteration of the course and students were asked to watch pre-recorded video lectures online, while class time used simplistic PowerPoint slides that framed the in-class discussions and activities; a ‘flipped’ teaching and learning model.
During his presentation, Prof. Nimijean displayed the final grade distribution of his two classes – the class that followed a traditional lecture driven approach plotted as a bell curve and the flipped approach plotted as an inverted bell curve.
So should we expect a normal distribution when plotting the outcomes of instruction? If the assessment you conduct is criterion-referenced and not norm-referenced, then we, instructional designers, would answer “no.” Educational psychologist Benjamin Bloom proposed that the “normal curve” (many students learning medially, few well and less well, some very well and very poorly) should not be an expected model of outcomes of instruction. This kind of distribution, Bloom argued, is what we might expect to occur without the intervention of instruction. Instruction should foster learning and scaffold learners. He argued that most students can master what we have to teach them, if we provide a learning environment conducive to mastering the subject (Bloom, B. S. (1968), Learning for Mastery). A body of research conducted since then, including a comprehensive, meta-analysis review, has confirmed the initial research by Bloom (Kulik, C. C., Kulik, J. A., & Bangert-Drowns, R. L. (1990a). Effectiveness of mastery learning programs: A meta-analysis).
Prof. Nimijean noted that he had more students perform very well after changing his approach towards more student-centered models of instruction, but he also observed more students perform very poorly, and the majority of students performing less than medially. The inverted grade distribution shows the impact of his new teaching strategy for a number of students who demonstrated learning outcomes well beyond the average and pleasantly surprised the instructor.
The case that the outcome distribution became inverted in the second version of the course is, in my opinion, the consequence of changing more than one variable in the new iteration of the course: besides including activities that support higher-level thinking and engagement (the student centered approaches), the course appears to have lost some the scaffolding that the majority of students may have needed (lecture notes, text in the slides, more structured class, etc.).
Providing additional material in a different format not only helps students who have visual/hearing challenges, but also the students who have a different learning style preference. There is also a significant amount of research that shows that minimally guided instruction is “less effective and less efficient than instructional approaches that place strong emphasis on guidance of the student learning process” (Kirschner, Sweller, Clark, “Why Minimal Guidance During Instruction Does Not Work”, 2006.). Prof. Nimijean realized this immediately and noted that the next time he would try the new approach with more “mature” students, which is also supported by this same research (“The advantage of guidance begins to recede only when learners have sufficiently high prior knowledge to provide “internal” guidance.”).
I think that Prof. Nimijean is on the right track by courageously trying out new teaching strategies. Going back to the “Goldilocks” analogy, I suggest keeping the best of both approaches – return the scaffolding that the majority of students need in terms of supporting material and structure, but keep activities that encourage higher participation and higher order thinking with the hopes that the final outcome distribution remains normal but with a smaller variance and a higher peak translated towards higher summative achievement scores.