Little evidence for the effectiveness of computer-based Personalized Instruction

Complete report

Related article

There has been a renewed interest in and enthusiasm for online learning and computerized instruction. One gets a sense of déjà vu when reading today’s educational blogs and policy documents, which are recycling the same arguments for computerized instruction that appeared in the 1980s. But in the more than 30 years since the personal computer and computer-assisted instruction entered K-12 education, not much has changed. Computers are now commonplace in the classroom, but teaching practices often look similar, as do learning outcomes.This raises two questions: What has changed to get people excited about online learning? And is this revival of enthusiasm warranted?

It seems that the pace of technological advancement,combined with the clear success stories of how technology has improved productivity in other sectors, is leading policymakers and educators alike to take another look at computers in the classroom, and even at computers instead of classrooms. In particular, advances in computational power, memory storage, and artificial intelligence are breathing new life into the promise that instruction can be tailored to the needs of each individual student, much like a one-on-one tutor.The term most often used by advocates for this approach is “Personalized Instruction.”

However, despite the advances in both hardware and software, recent studies show little evidence for the effectiveness of this form of Personalized Instruction.This is due in large part to the incredible diversity of systems that are lumped together under the label of Personalized Instruction.Combining such disparate systems into one group has made it nearly impossible to make reasonable claims one way or the other. 

To further cloud the issue, there are several ways that these systems can be implemented in the classroom. We are just beginning to experiment with and evaluate different implementation models—and the data show that implementation models matter.How a system is integrated into classroom routines and structures strongly mediates the outcomes for students.

In light of recent findings, it may be that we need to turn to new ways of conceptualizing the role of technologyin the classroom—conceptualizations that do not assume the computer will provide direct instruction tostudents, but instead will serve to createnew opportunities for both learning andteaching.