Science Magazine published an RCT of the Modeling Practices in Calculus (MPC) approach to teaching college calculus. Quick take: Sizable positive impact on end-of-semester test scores, but findings aren't definitive due to the key study limitation of faculty self-selecting into teaching MPC.
Program:
-
Per the study: "The MPC approach is designed to bring the authentic practices of mathematicians into the classroom, by facilitating active student engagement in the practices of mathematicians to learn calculus in a student-centered environment."
Study Design & Findings:
-
The study randomly assigned 1,019 students at Florida International University who'd enrolled in calculus to classes using MPC (treatment) vs usual instruction (control). The study found sizable, statistically significant impacts on the end-of-semester test (0.77 standard deviations).
-
But I believe a key study weakness is that participating faculty weren't randomly assigned to teach MPC vs control. Instead, 7 faculty volunteered to teach MPC (including its two-day training and weekly preparation meetings); 12 faculty stuck to their usual teaching methods.
-
The act of volunteering for MPC may indicate a level of professor motivation and/or capability that's higher, on average, than that of professors who didn't volunteer. This difference in faculty motivation/capability (rather than MPC) could potentially account for the superior outcomes for the T group.
Comment:
-
Bottom line: I think the findings are encouraging (due to their magnitude) but less than definitive. This would be an excellent candidate for a replication RCT that recruits faculty willing to be randomly assigned to the new MPC approach vs their usual teaching methods.