Adaptive learning, or a few words about Knewton

    If you are interested in modern technology in education, then you probably already know about Knewton. If this is not the case, then the information below will be useful to you!

    Why is it important?

    Knewton is known for being one of the first to actively apply data analysis technologies in education. As a result of this work, an adaptive educational platform was created that can be connected to any modern educational process management system (LMS).

    The idea that the educational application adapts to the student’s unique “learning curve” has great potential. Moreover, the stack of data analysis technologies that allow you to build such a system is in a fairly mature stage. Despite this, such technologies remain closed to most players in the educational market due to the high cost of development. Only organizations with large resources can afford such a system, for example, Khan Academy, which received multimillion-dollar grants from The Gates Foundation, Google and others.

    A ready-made platform that allows any educational institution to implement personalized learning is a big step forward in the field of educational technology development.
    As Charlie Harrington, head of the London Knewton office, writes:
    Imagine that a teacher can, with a couple of mouse clicks, evaluate individual student knowledge in his subject at any given time. This will help teachers easily and quickly identify topics in which a knowledge gap is just beginning to arise and change the learning process in such a way as to fill this gap. Teachers will have more time to do what is best for them - to inspire and teach.

    Charlie Harrington speaks at the EdCrunch conference in Moscow

    Adaptive Learning

    The idea of ​​adaptive and personalized learning arose in the 1950s and goes back to the “teaching machines” of psychologist B.F. Skinner, the founder of behaviorism, was then a professor at Harvard University. Based on the principles of learning that he developed during experiments with pigeons, Skinner created a mechanical device resembling a box that “fed” students with questions. Correct answers were rewarded with new academic material; wrong - led to a repetition of the old question. “The student quickly learned to answer correctly,” said Skinner.
    "Learning Machine" by B.F. Skinner

    The movement became popular in the 70s on the wave of interest in artificial intelligence technologies. Then scientists believed that sooner or later a computer will be able to adapt to the environment no worse than a person. The use of machine learning mechanisms in education has become a fashionable topic in the scientific community, however, the cost and size of computers of that time deprived this undertaking of any practical meaning.

    Only by the end of the 2000s did the idea begin to take on real shape, and adaptive learning became fashionable again. Systems like Knewton today have a wide range of features, such as sophisticated skills development tracking, instant feedback, personalized prompts, and what was not available to Skinner's Harvard students - an interface that resembles a computer game!
    Knewton Interface

    How Knewton Works Knewton

    founder Jose Ferreira spent his whole life engaged in educational technology. Since 1991, he worked at Kaplan, one of the largest players in the market for paid educational services. In 1993, he tried to bring the idea of ​​adaptive learning to the company, but could not move the slow-moving corporation from its place, which was not surprising in general - in 1993 only a few people had computers! Jose was ahead of his time, and when technology reached the right level of development, he founded Knewton in 2008.

    Knewton founder and CEO Jose Ferreira

    The Knewton methodology is built around two basic concepts: technology for planning an educational trajectory and a complex student assessment model. This approach is fundamentally different from most “adaptive applications”, which essentially apply the adaptive approach to the only point at which students' knowledge is measured. An example of such a “poorly adaptive” approach is a diagnostic exam, based on which the computer determines what content will be shown to the student in the future. Data mining and personalization technologies are used here minimally or not at all.

    Adaptive learning in the understanding of Knewton must respond in real time to the results of an individual student and his actions in the system. This approach increases the likelihood that the student will receive the right educational content at the right time and achieve their goals. For example, if a student does not cope well with a certain set of questions, then Knewton will be able to guess which topics covered in this list of questions turned out to be incomprehensible and offer him content that will help increase his understanding of precisely these topics.
    The individual educational trajectories of two students

    Knewton calls himself an additional layer of an educational application that analyzes data. That is why any educational institution or project can work with Knewton. The data that the adaptive platform uses is collected by the educational application itself and transmitted to the Knewton server using the API. To start collecting a certain type of data, for example, when a student begins to watch a video or the result of answering a question, just add one line of code that will transmit this data to Knewton. The adaptive platform analyzes the collected data and returns it to the application in the form of recommendations to the teacher or an indication of which block of content should be shown to the student next.

    Arizona Dream

    Arizona State University is the largest state university in the United States by the number of enrolled students (annual enrollment is 70,000 people). Its president, Michael Crowe, a rebel and troublemaker in the academic world, called his school “the new American university” and chose a strategy for actively introducing modern technologies in the field of e-learning. It was within the walls of the University of Arizona in the fall semester of 2011 that the experiment on the introduction of adaptive learning began, in which Knewton participated with his partner, Pearson, a giant in the world of paid educational services.

    Role of Knewton and Pearson in the Arizona University project

    What was the experiment? It was decided to introduce an adative system for preparing first-year students in mathematics. The adaptive learning system had a double focus and, on the one hand, helped teachers, on the other hand, helped the student in autonomous work on the material. She used the data to understand the student’s level of knowledge and which teaching method is most effective for him. Based on the analysis of these data, the system made a recommendation on the sequence of study topics. On the other hand, Knewton provided instructors with real-time reports that helped them identify weaknesses in student preparation, create an adapted curriculum for everyone, and pay special attention to the topics that students learned the worst.

    Preliminary results of the experiment showed that the results improved by 18%, and the percentage of deductions fell by 47%. These results inspired the Gates Foundation in 2013 to launch a special program to accelerate the spread of adaptive learning technologies. The experiment also had a significant economic result: this program helped the University of Arizona earn an additional $ 12 million in additional tuition fees. As Joseph Ferreira noted in an interview, the University of Arizona pays Knewton $ 150 for a student who uses an adaptive platform. Critical Knewton

    Before and After

    Despite the huge and obvious successes of using adaptive learning, there are still many skeptics and critics of this approach. The first "stone" in the garden Knewton throw pedagogy. They argue that the Knewton approach comes from the fact that there is always the right answer. Is this approach universal? And will not the availability of tools that work well for the exact sciences impose this model on other disciplines?

    Another issue that often arises in connection with Knewton is personal data. Knewton claims to not store student personal data. According to Jose Ferreira: “We help the student understand his educational history without preserving his identification information.” Despite this, questions continue to arise. For example, according to the famous blogger and journalist in the field of educational technology Audrey Watters, “What does personalization mean if we cannot identify a person?”

    Knewton in Russia

    Taking the opportunity to make an announcement - December 1 at 19:30 at Digital October Knewton conducts an open lesson “Analyze this: how big data will revolutionize education”. Speaking will be the head of the London office of Knewton Charlie Harrington. An open lesson will be held as part of the online courses producer program at the New Professions Lab .
    Also - a video of the speech of Charlie Harrington at the EdCrunch conference on November 18 in Moscow.

    Also popular now: