Learntech is a tech application, that supports learning. We all learn. You can be a kid or an adult, employed or unemployed you can learn at school or home, you can learn in a group or alone… It doesn’t matter, learning or personal development are everywhere and can take place at any time. Which are the challenges for learntech designers and users?
Where the purchasing power is, there is influence…
Designing edtech products for and pleasing diverse groups—students, teachers, administrators, and parents—is no easy feat. This forms the first challenge for learntech, but directly follows the next challenge. Among these groups, the administrators are often the ones with the most purchasing power, and students are typically the ones with the least. That can lead developers to prioritize designing tools for administrators’ experience, even if the product is meant to be used in the hands of students and teachers.
Still there where is the purchasing power, there is influence… To convince the efficiency or impact of an edtech tool numbers matter!
I can see 3 main subcategories in learntech tools:
- Tools for content management
- Channels for content
- Supporting the learning process by using data as a source to measure. In fact, we could measure all different learning actions and store this in a controlled way.
“Everything that can be measured is measured. Everything that is not really measurable is being prepared to be made measurable … number mania ”
The next learntech challege is how can we measure the learning process? Educators want to see measurable growth. Both educators and edtech executives are aligned on this question: How much better will students do this year compared to last year with the support of this product?
That’s why is not at all bad for edtech designers to go again to the school banks on one side of the other… to observe and learn how their product or service may influence and support the learning of teaching process. It won’t take long in the classroom to realize the flaw in this upside-down logic. It’ll become apparent that the most important consideration in designing edtech tools should be students’ joy, engagement and learning.
To support the learning process edtech tools have to give challenges to the students. If they only give the output wright or wrong and even the correct answer it will not help the learning process. If your tool does only this, it is based on the old teacher-centered way of learning. They will learn by heart today, but has forgotten it tomorrow…
Challenges for the teachers or learners.
As a teacher or student, it is not always easy to find the right tool. Ask yourself what is the purpose of using the tool. All curriculums are written by somebody and there are always biases in it. Most of the time we don’t consider to realize that the values of the writer/ designer are incorporated in the tool.
You always have to ask yourself who designed this? Who wrote this for whom? And this is not only for technology but also for books and articles ( yes check my background you will learn about my “glasses”!!!)
Some questions to consider :
- Who is writing the algorithms?
- Who is designing edtech?
- Who designs AI for learning experiences?
Does AI has a place in the learning process?
How does AI meet the challenges of the designers and users?
1. Make learning/ teaching easier
Technology is an application, so it has to make something easier not more complicated! You have to use the tools properly. Be aware technology support learning or teaching but it will also support bad learning or teaching. If you don’t know how to use it, ask a digital expert or the edtech startup.
But the first step is to ask yourself why you would like to use this tool? What is the purpose of using it?
2. Measurement of the learning process
Measurement is possible with the help of an Application Programming Interface (APIs). An API defines the access to the functionality that lies behind it. The outside world knows no details about the functionality or implementation, but thanks to the API it can use that functionality. An advantage is that multiple implementations can be approachable with one API, as long as they meet the API. The API captures data from different technologies and record the data in an LRS (Learning Record Systems) in the form of noun, verb, object. A example is
The different LRS systems can communicate with each other via an API so that a wide range of data and experiences (also from the real world) can be stored. This can also be integrated with LAP (Learning Analytic Platforms ) dashboards with learning analytics which can, in turn, be connected with LMS ( Learning Management Systems from the school or company). Remember who has the purchasing power!
For the individual learner, this means that each learner can have his data locker (box) in which his / her personal learning information is contained.
AI technology within Edtech tools can be used to close the inequity instead of enlarge inequity. AI recognize patterns in data. These patterns can provide information to teachers and can help support better teaching. In this way, AI edtech tools can provide the information to the teachers so they understand the patterns and use this information in their classroom.
Bias in algorithms
Diversity in the design team is important and also applies diversity in data. Be careful that the trained algorithms are trained by diverse target groups.
Computational thinking, programming will not help the ethical question. What questions do designers have to ask towards schools:
- Who is not served by the recurrent system?
- Who are you designing for?
- What are their goals?
- Which approaches do help them?
- In which way can technology support them?
- What are we trying to achieve? It has to fit with the learning goals!
- Do stakeholders have the chance to give feedback on the algorithms?
You need metrics to measure the results it is difficult because there is a lot off noise. It is really hard to prove the real support. You need also to be aware of the different subgroups and cross-cultural influences. Cultural misunderstanding can lead to new insights and ideas… For example the effect of feedback on behavior. Negative feedback gave positive behavior with Chinese and positive feedback give positive behavior for Americans.
There has to be a lot more progress in algorithms and AI before we’re going to have a system that could show you the steps and know why you did it wrong. To get to human-centered AI, we need to have new algorithms that are going to push us further. So deep learning is not going to do it.