In my last blog I wrote about how the traditional school system hampers student progress by tethering all students to common pacing, thereby handicapping a student like Harrison Bergeron from achieving his full potential. Rather than allowing students to progress as they master concepts, the traditional system forces students to accelerate or pull back to keep pace with the teacher’s batch process at the front of the class.
Recently Dan Sturtevant from the Massachusetts Institute of Technology conducted a novel study that uses simulation modeling to compare the traditional model, in which students progress at a common, regulated pace, to a network-centric model, in which students work at individual paces and advance as they master material. Sturtevant’s research brings the science of system dynamics to the field of education. By studying the underlying architecture of the system, Sturtevant quantifies the theoretical losses that result from the current system design.
Sturtevant describes the traditional classroom as having a value-chain structure. Students in the value chain progress through a series of discrete subjects and grades in lockstep with each other, and each classroom generally operates as an autonomous modular unit, as depicted below:
He compares this system to one in which the Internet allows for dynamically responsive learning. Students self pace based on mastery of material, and teachers act as tutors to help each student past sticking points. This network-enabled paradigm eliminates cellular pacing requirements and allows for students to move forward at individually maximized rates.
In his first simulation, Sturtevant assumes, for the sake of experiment, that student ability levels do not vary. He wants to measure the effect of common pacing alone. The simulation results show that due to common pacing, none of the students obtains the 16 possible units of knowledge available in his model. Most achieve about 13 units. Furthermore, about 20% of the students fall off significantly, achieving as few as 9 units. Remember, this fall off is due to the pacing penalty alone, not variation in student ability.
Such simulations are only theoretical and do not map directly with reality. However, they are useful for understanding the potential long-term trends that result from various controlled scenarios. In Sturtevant’s simulations, the implication is clear: value-chain architecture leads to quantifiable and unavoidable loss in student learning.
Sturtevant’s work brings some data-driven backbone to my rant last week about the costs of the traditional, group-paced system. Not only does he thereby neatly demonstrate how MIT can always lend a helping hand to pick up where well-intentioned Harvard left off, but he also introduces a useful way to use simulation modeling to advance education systems design.