In a typical learning course, students will have objectives or criteria they must meet by the end of the course. Students will study content through a learning process so they can demonstrate they have met these learning objectives, normally through a mix of formative or summative assessments. While all students must meet these objectives by the end of the course, how students get to that point can vary and be adapted to the student’s needs. This is the premise for differentiated learning; the content, learning process, and even the end-product can all vary and adapt to a student’s strengths and maximize their learning potential.

Implementing differentiated learning into a learning course has many advantages, such as personalized content that focuses on a student’s weaknesses, using a student’s favored modes of learning more frequently, providing personalized feedback based on a student’s solution, among many others. These advantages have an even larger impact when used in e-learning ecologies. The technologies available allow for a wider range of adaption and personalization that can happen much more frequently compared to the capabilities of a single teacher in a classroom. So, its not a question of if differentiated learning should be used in e-learning ecologies but is a question of how differentiated learning should be used.

There are several different types of differentiation as summarized by Kathleen Scalise [1]. Boolean or model-based differentiation is what we would commonly associate with adaptive learning environments. A system is designed to change the content or process a student works through based on inputs from the student. On the other hand, self-directed approaches allow students to choose their own learning path, with little or no control from the system. While both approaches give students different amounts of control over their own learning, both lead to a learning path that differs between each student. So, which approach to differentiation would be most effective in e-learning ecologies?

The Boolean or model-based approach would be designed based on pedagogical theories and big data. Such a system would create a learning approach for a student by comparing that student’s skills and weaknesses to others like him or her. While the designed approach would meet the required learning objectives of the course and have some amount of personalization, the system may lack the fine-tuning needed to differentiate the student from the other students on whom the system was based upon. This does depend on the specific system, and with advancements in AI, may not be an issue in the future.

Self-directed approaches offer a freer learning experience for students where each student can choose their next step along their learning path. This obviously offers a higher level of customizability, but risks students going down the wrong path, potentially leading to a more time-inefficient learning process or even missing content for some learning objectives.

Like a lot of things, I think the best approach is to use a healthy mix of Boolean/model-based differentiation and self-direction approaches. Students should be given some control over their learning path, to boost engagement and to develop their independence, but a system should also be in place to make sure students are learning the correct content in the most efficient manner for their specific needs.

How much control do you think students should have over their own learning path?

[1] Kathleen Scalise (2007) Differentiated e-Learning: Five Approaches through Instructional Technology