If we personalize learning, do we learn better?

**MOOCs (Massive Open Online Courses) are large-scale, open distance-learning courses delivered online to an unlimited number of students. Learners can access traditional course materials such as readings, videos and quizzes, with additional support from online tutors. MOOCs have revolutionized distance education—but have they also revolutionized learning? Learning needs to be adaptive and personalized. Otherwise, the results may not be positive.**

In this TEDx Barcelona talk, Iván Ostrowicz, co-founder of Domoscio, a company specializing in adaptive and personalized learning, explains that everyone learns differently, making individualized, personalized education ideal.

Iván Ostrowicz holds engineering degrees from the Polytechnic University of Catalonia and ENSIMAG, and an MBA from Neoma Business School. He has international experience in information systems and organizational management at major companies including Air France and EY. He is the co-founder of Domoscio, a company specializing in adaptive and personalized learning.

Watch the video

TEDx Barcelona

Read the video transcript

According to Iván Ostrowicz

Since we are going to talk about education, training and these kinds of things, let's start with some participation. I'd like you to raise your hand if you think we all learn in the same way.

Raise your hand. Whenever I finish this exercise, I have around 23 people. I have done it all over Europe, always with the same result.

It is a fact that we all learn differently. The reality is that we have an education system with one teacher for 30 students. At university, it is one teacher for 200. If we think about continuing education, or many modern learning systems, we have one teacher using a single teaching approach, a single methodology for 10,000, 20,000 or 100,000 participants.

But let's go a little further and see whether you remember something we learned at school. Let's complete this sentence together: I learn, you learn, he or she learns, all of us learn.

Back in the late nineteenth century, Hermann Ebbinghaus, considered the father of experimental psychology in education, conducted an experiment demonstrating that after a month, if we do not study or review, we forget 80% of what we have learned.

So we face two facts: We all learn differently, and we all forget what we learn. This raises the challenges of how to personalize learning and how to consolidate what we have learned so that we remember it.

Part of the answer can be found in cognitive science

In 2010, Ming Zher Poh, from Harvard and MIT, conducted an experiment using a device connected to a person that made it possible to observe brain activity during different activities. As you can see, brain activity in class was the same as when watching television. My apologies to the teachers here—it was only with one particular teacher.

Similarly, when people were in the laboratory experimenting, there was much more activity. When they were doing exercises, there was also much more activity.

In fact, in 2006, Rodríguez conducted an experiment to identify a good study method. He divided people into groups who had to learn from a text:

The first group was allowed to read the text four times.

The second group read the text once and completed a set of exercises three times.

After a month, he found that people who had done exercises remembered 52% more of what they had learned and, above all, had spent four times less time studying.

I could go on to talk about the Leitner system. It is a flashcard system with a question on one side and an answer on the other. You move the cards farther back in a box as you learn. We could also discuss learning speeds, learning preferences and different forms of intelligence.

Another part of the answer may be found in new technologies

If we talk about new technologies in education today, we are talking about big data. What is big data? We generate data all the time: when we use our phones, when we use a platform and when we are online. Today, storing all this is very easy. It is inexpensive, and we can easily access the information.

Think about uploading something to Dropbox, where different systems copy it. Big data also gives us computing power that did not previously exist, and that is inexpensive and easy to access. This makes the artificial intelligence algorithms developed in the 1980s and 1990s fully relevant again, allowing us to do things we could not do before. So what happens if we combine cognitive science with big data and artificial intelligence?

Adaptive and personalized learning

In 2013, IBM identified the combination of these three fields among the five technologies that would emerge and revolutionize education. Only three years have passed, and it is called adaptive and personalized learning. It involves analyzing everything students do on a platform: exercise results, the path they follow, whether they consult the help section, and whether they achieve better results by watching a video, then doing an exercise and then reading a text. Combined with different elements, this makes it possible to offer a learning experience experimentally adapted to each person—their level, pace and preferences.

We have already done this. We worked with a French publisher on mathematics, French and English content for 10-year-old children. The publisher provided content adapted to work with these algorithms, and we put it on a platform: an app on iPads connected to our algorithms.

We analyzed everything students did on the platform. With 400 students, 87% of our recommendations were successful and helped the student reach the end of the process—in other words, learn something. When students got stuck because content was missing, was more difficult, or for another reason, teachers had learning analytics showing where each student stood. This enabled a human intervention.

As you can see in the second example, after students had learned something, we asked a question to check whether they had really understood the concept. Based on that question, we calculated when they were likely to forget it and made studying easier. We then sent them a question on the same concept to refresh their memory. The question always changed, because we did not want them to remember the question instead of the concept behind it.

Another project involved a client in a completely different situation: training managers to be good leaders. Training took place in person with a teacher and included role-playing exercises on how a manager should behave in different situations. After the training day, we sent a set of questions, one for each concept. This told us what they had and had not understood.

From there, we again created a review schedule, sending a question about the right concept so they would not forget it. We presented it just before they were likely to forget. This also demonstrates that the time it takes to forget something is different for each person and each concept. Perhaps we love mathematics and find it easier to remember, but if we dislike history, we will have to work harder.

Results of adaptive and personalized learning

After an average of three months, people who completed this training remembered 79% of what they had learned. That is 8 concepts out of 10. The person who used the system the most reached 90%, while the person who used it the least remembered only 5%.

Returning to MOOCs, what is the problem? Today, dropout rates range from 75% to 95%. Only 5% to 25% finish a MOOC.

On the other hand, MOOC platforms have large numbers of participants generating a great deal of data. We are now working to include our technology in MOOCs. We will consider more than each person's path, exercises and answers, or whether they prefer videos or achieve better results by reading a text.

We will assess users' responses in the forum. If a participant does not understand a concept and asks a question, and another person answers it well according to the community, that person's score increases.

We will take that score into account. If a person knows a concept very well, perhaps they do not need to complete that chapter. Perhaps they should move through it faster, or be offered a more difficult level. Let's have a little more participation: when I was a student, these were the kinds of solutions I hoped for.

Many things are changing in education today. MOOCs, for example, have improved access to training, but that does not mean education itself has improved. If we consider that we now need to learn throughout our lives, this kind of solution becomes even more important.

Big data for learning exists, and it is called adaptive and personalized learning. It is breaking down teaching models worldwide and opening up a vision of lifelong personal development. Ultimately, adaptive and personalized learning helps us learn better and remember longer.

Thank you very much!