For schools and universities to implement this learning style across the curriculum and across all courses, an unprecedented educational paradigm shift is required. In the history of education, this has never really been attempted, except perhaps outside of a handful of specialized schools and universities. Although adaptive learning is a departure from traditional pedagogical methods, educators hope that this will become the new norm, over time, with courses tailored to the unique needs of each student.
What is adaptive learning
Adaptive learning is a technique for providing personalized learning, which aims to provide efficient, effective and personalized learning paths to engage each student. Adaptive learning systems use a data-driven approach to adjust the path and pace of learning, enabling the delivery of personalized learning at scale. Also, adaptive systems can support changes in the role of teachers, enable innovative teaching practices, and incorporate a variety of content formats to support students according to their learning needs.
Knowledge is a graph
The strongest predictor of how we perform in a learning environment is our prior knowledge. What we already know before starting the learning experience. A 1999 psychology article by Dochy, Segers, and Buehl found that prior knowledge is 81% of the differences in outcomes between students. Reviewing prior knowledge before showing new information affects learning outcomes. And connecting new knowledge with prior knowledge while teaching can also have a huge impact.
The most famous psychology article is about it, it dates from the year 1956 “The magic number seven, plus or minus two” by George Miller. The paper suggests that humans have limited working memory. Miller discovered that for simple numbers, a human could work with approximately seven elements at a time. Later researchers found more complex information, that limit is closer to four. In short, we learn by connecting previous knowledge with new information. And those connections form a great, endless graph of knowledge.
How do we know if adaptive learning is good?
Since these systems come from academia, we have a significant amount of data and history with each system. Human individual mentoring has the strongest learning outcomes. This is a common finding in educational research. Until now, no computerized adaptive learning system has surpassed individualized human tutoring.
Some important adaptive learning systems
Some Important Adaptive Learning Systems One of the first implementations was the Skinner teaching machine. During the 1960s and 1970s, there were several attempts at computerized instructional systems. Costs and slower machines limited the success of these systems. In the late 1970s and early 1980s, Item Response Theory allowed test manufacturers to begin work on computerized adaptive testing. An early and influential computerized system was the Lisp tutor, also known as LISPITS (1983) at Carnegie Mellon University. SuperMemo, launched in 1985, incorporated spaced learning into a computerized system.
Also in 1985 appeared the role of Knowledge Spaces, which forms the basis of one of the four elements. The ALEKS math tutor emerged in 1994, strongly promoting the use of knowledge spaces. In 1995, Corbett and Anderson published “Knowledge Tracking”, which forms the basis for Bayesian knowledge tracking models. Some important programs include AutoTutor, ACT-R, and Cognitive Tutor Authoring Tools. Knewton is an example of contemporary adaptive learning systems. Kaplan and Pearson use Knewton to provide adaptive learning experiences.
The four elements
Most of today’s adaptive learning systems have these four elements. The terms change and so does its scope. But almost always the following will be found: The expert – a graphical model of the “ideal” state, of everything that the person could learn using this system.
Students: a model of the student’s current state, showing the probability that the student knows each of the nodes in the expert graph.
Tutors: the algorithms that determine what content to show and when.
Experts models and learning models. The learning model inform the tutor. The tutor seeks to optimize the content for its relevance, challenge, and efficiency.
Interface: which is how to show the learning experience to the student. In many adaptive learning experiences, the interface changes based on the student’s model and the tutor’s goals. Some psychologists suggest that of these “four spaces”, at least one or two must be prior knowledge. The amount of prior knowledge that we can “load” into one of the four slots depends on the strength of the connections on the graph. When we have previous and new knowledge in our working memory, we associate the information. And we strengthen the connection between the two. Trying to learn new information without connecting with prior knowledge limits the strength of memory.
Benefits of adaptive learning
Because each student learns in different ways, even when classified into broad groups (eg visual, spatial, logical, social, etc.), students will have different learning outcomes. Not everyone is going to absorb the knowledge that a teacher is trying to provide in the same way, so some students will understand it, while others will struggle. Although some of these outcomes are the result of individual characteristics, intelligence levels, and any known learning difficulties (for example, students with ADHD or dyslexia), educators are responsible for how a course is delivered, and therefore for the outcomes. results to expect.
Maintaining a one-size-fits-all approach doesn’t fit the needs of modern students as well. Students of all age groups are more immersed than ever in digital ways of learning and thinking. When the concept of adaptive learning first appeared, that’s when computers began to become mainstream. It was envisioned that AI programs would tailor courses to the needs of individual students. One system that emerged at that time was known as Scholar, which laid the groundwork for future adaptive learning attempts.
Improvement areas
Researchers have investigated classroom learning alone, computerized adaptive learning alone, as well as blended learning and adaptive learning. Adaptive learning systems generally outperform traditional learning in the classroom. Combined with learning in the classroom, adaptive learning systems create a positive effect, but there are some limitations. Adaptive systems work particularly well with instant feedback and ensure mastery of skills. The researchers point to some areas for improvement: The cost of developing content for these systems is high. These systems often cannot contextualize learning in the way that a human being does. Adaptive learning systems can seem more challenging, which can reduce student motivation.
Adaptive Learning and Predictive Analysis
The combination of adaptive learning with predictive analytics has great potential to improve the way students learn and generate positive learning outcomes. AI-based learning systems can collect and process large amounts of data from student learning activities, such as the amount of time spent completing each task, response latency, and assessment results. The data can be used to detect patterns and build predictive models that help identify individual student needs and improve the content delivered to each student. Algorithms analyze data much faster than humans. Thus, students are provided with content, prompts, and interventions, all of which change in real time based on their individual needs and abilities. Although many educators can see the benefits of adaptive learning, the challenge is finding a way to implement it and do it cost-effectively.
Adaptive learning in practice
Fortunately, we are now at a point where educational software is advanced enough that it can be more easily adapted or customized based on the needs of students, educators, and content creators. Rather than offering a single learning package for a course, educational content creators and providers can tailor learning packages to a variety of needs. A quick assessment, which any teacher or tutor can implement, should determine the learning styles present in any particular course and class. With that information in hand, teachers can use educational management software to implement a series of options for different learning styles in each class. Students can receive a number of options, from traditional instructor-led teaching to interactions with videos, quizzes, activities, learning sessions, and software programs on a tablet, phone, or computer.
Conclusion
Adaptive learning is the future of education. Sooner or later, students around the world will benefit from being able to select courses and modules that are more closely tailored to how they prefer and need to learn. Schools and colleges that offer adaptive courses, with the software to deliver them, will gain an advantage over those that do not. Despite cost concerns, with the right resources in place, creating a series of learning pathways costs no more than preparing traditional instructional materials and forms.
Teachers also need not worry about adaptive learning taking over the core elements of a course. Instead, adaptive elements can be elective or basic learning can be taught first, followed by a period of a lesson in which different students approach study in a variety of ways depending on what they need and how they learn. However, it should be noted that adaptive learning may not always easily connect to all courses, disciplines, and subject areas. Educators must always make a decision, taking into account in part the needs and styles of the students expecting a new entry, along with the demands of the course and anticipated learning outcomes.
Bibliographic References
Beltrán Llera, J. A. (2003). Estrategias para una innovación educativa con Internet. En Fundación Encuentro La novedad pedagógica de Internet. Madrid, Educared.
Gallego, D. J. Y Alonso, C. M. (2002).Tecnologías de la Información y la Comunicación. Madrid, UNED.
Majó, J. Y Marqués, P. (2002): La revolución educativa en la era internet. Barcelona, Praxis.
Adaptive learning
