Why Technology Can’t Replace the Person
By Matthew S. Howell
However he envisioned the state of education and learning, he certainly left room for a world where the concept of personalized learning could take root and flourish. Personalization through technology is a response to scalability issues and a means to reach curricular goals through variations in scope, pacing and sequence. The algorithms used in this process are similar to those used when looking for a show or movie on Netflix. A student is given a set of possible options which unfold based on data norms of similarly performing students with a match in learning profiles. This means that the personalized technology is merely responsive, not adaptive, and not truly personalized. The options are selections from a menu.
As is implied in the word personalization, it is done best when a person is doing it. As all good teachers and educational leaders know, there is nothing more powerful than when a person works to meet the needs of another. These types of interactions are the lifeblood of education and the central focus of any successful school.
In order for this to happen through technology, the technology must become intuitive and adaptive, not merely responsive. This means that the technology needs to “think” and behave the way that a person does, or at a very minimum, it needs to use facial recognition software to understand a person’s emotional state. Technology could then conceivably fabricate approximated emotional responses and inspire some level of critical thought and questioning.
There are numerous programs in the educational marketplace which are billed as personalized and individualized. While these claims are unchallenged for the most part, there has been little analysis done to quantify if these programs actually deliver. The one area where there is little question of success, is that these systems are superb for gathering and storing student data. The bold claim is that this information will target learner need, but the jury is out, and there are very few independent studies which show that this is a successful model.
In spite of decades of standardized assessments the feedback loop between assessment data and well informed instructional practice is fuzzy at best. Given that many personalized learning tools are based on wholesale data collection and analysis, one must ask that question; is the current trend of personalized learning just another step toward learner standardization?
In a truly personalized model, one would argue that learner autonomy and interest are primary tied to desired outcomes. At the center of such a model, one would find a learner who is excited to pursue a passion and guided rightly to reach the desired objective. The emotional intelligence of the individual would be the most valued learning goal and the learner would understand himself and pursue knowledge aligned to his intrinsic fulfillment. Studies have shown that the psychological trend for students using current personalized technology is often counter to this outcome. Instead, the student works to improve his score in a given area, based on the menu options. The learner begins to quantify his success by improvements in his data not by a guiding principle of fulfillment.
The difference between a data collection model and an individual pursuit model boils down to the difference between compliance and exploration. A responsive model of software-driven learning tools eliminates learning options and funnels a student’s inquiry through a predictable labyrinth toward a final predetermined goal.
If personalized learning is designed to bring a learner to an end success, what is that success? Given that the algorithms have a predetermined end point, is the final goal built and aligned to the definition of success as measured by student and teacher input? Have parents and students been consulted on their definition of success, or is it a foregone conclusion that exemplary data is the definition of teacher and student success? More often than not, the personalized learning program is created to meet a large-scale curricular goal aligned to standardized outcomes. While the drum beats steadily in this direction, has consideration been given to the impact on the student?
There are those undaunted champions of this phase of technology development who rightly point out that there are programs which appear to be more adaptive. One needs look no further than Smart Sparrow and Knewton to find examples, where the promise of adaptive learning looms large. These programs are using machine learning to adapt to student proficiency. In so doing, there is expansion away from the decision tree model. There is the promise of cognitive tutoring and timely feedback, which is standard in the industry.
Consideration should be given about what this means for the students who are tasked with using these programs on a regular basis. Feedback is only of value when it is constructive and aligned to the attainment of a desired goal. The human brain is a goal seeking organ, and it will develop neuropathways directly related to an individual’s experiences. Educators are the gatekeepers of such information.
Author