Adapting Adaptive Technology

A wide variety of educational reforms and programs have been championed in a pursuit to improve education in the United States.  Some recent ideas include enforcing standards, opening charter schools, providing vouchers for private education, improving teacher pay; two recent federal programs include No Child Left Behind and Race to the Top.  Despite these efforts, by almost any standard the quality of education in this country is lagging, with U.S. schools still languishing in the middle of international rankings, behind the schools of such countries as Estonia and Slovenia (Mehta, 2013).

Learnatric believes there exists significant potential in utilizing adaptive learning principles into the educational sphere to a far greater degree.  In fact, studies have shown that tremendous impact can be created in this way (Adkins, 2009):

“Cognitive and intelligent tutors are meta-cognition technologies that simulate the behavior of a human mentor and provide personalized responses, remediation, and interventions in real time based on the knowledge, behavior, and cognitive abilities of a particular user.

The products are based on artificial intelligence and generate a cognitive model of the student based on the interaction with the student.  The model is then used to provide individualized instruction to the student.

In a seminal study known as the “Two Sigma Problem”, Bloom found that, on average, tutored students scored better than 98% of classroom students.  This means that the achievement of individually-tutored students may exceed that of classroom students by as much as two standard deviations (a two sigma shift).  This knowledge-transfer improvement is roughly equivalent to raising the performance of 50th percentile students to that of 98th percentile students.  New cognitive tutor products are capable of exceeding the two-sigma deviation.”

As Learnatric embarks on deploying its dynamic learning platform, we believe we can significantly impact education in this country in two ways.  First, we have the potential to deliver a more responsive educational approach to students – one that promises to support greater educational achievement in large numbers of students.  Second, we also have the opportunity to reshape theories in pedagogy and learning science.  Through our use of machine learning, we will have a “big data” component to the system that can monitor learning trends among large numbers of students.  By analyzing the data, we have the potential to validate or challenge existing theories about optimal educational approaches.

Leave a Comment