- Learning and Development
- Digital Workplace
- October 25, 2017
Personalized learning and how to effectively enhance it with modern technology
Let's first define "personalized learning". The term personalized learning refers to a set of educational methods and techniques for adaptation of a learning process to an individual's learning style, personality, unique needs, background and previous experiences.
Everyone's learning is based on previous experiences. As research shows, there is a significant improvement in a learner's achievement when learning is built on a platform of previous knowledge. In their bestselling book "Mindful Learning: 101 Proven Strategies for Student and Teacher Success", Linda and Bruce Campbell show how important a role is the prior knowledge that learners have. So, when one's previous experiences are known, learning could become more effective if adapted and linked with something that person already know.
Consider this example. Recently my son asked me "what is this and what do you need it for?" about the gearshift in my manual transmission car. After a short thought, I explained to him by referring to his bike's shifter and gear mechanisms, which allow him to change speed/effort ratio. By knowing something from my son's previous experience and linking new information to that, I made this new concept easy for him to learn.
What if I would be explaining it straight from the mechanic's point of view or drawing pictures with an engine, gearbox and all that stuff? He would probably lose interest and give up after the first half a minute as he has no ground that all that could be based on. As a result, no learning would happen.
To summarize this example, if a person's previous experience is known and a new concept could be linked to it, then much better results will be achieved in terms of understanding and learning.
Personalized learning in an organization
What if we imagine a similar situation in a company setup – there is an employee who needs to learn about the new tool, technology or process. If the information is linked to the person's previous experience, it speeds up learning and, as a result, improves the employee's performance. This employee is like my son in the example above. But in my example, I was the one who knew about my son's context and experiences, so, I was able to make a link between the two – the known and the yet unknown – for the unknown to be easier to learn. Is there anyone around in the father role who knows well the employee's context, can talk his language and knows about the employee's previous experiences? I bet, in most cases, there isn't such a person. So, how could learning in the company's environment be improved so that the employee will not lose interest and give up just because the new concept is too complex to catch at once and isn't linked to anything he already experienced?
Problems of traditional methods
One way to substitute the parent's role in a company setup is mentoring. The more experienced employee is assigned as an adviser to the less experienced employee. With all knowledge that the mentor gathered by experiencing something in the past, he could understand the issues and difficulties that the mentee is facing and guide the learning towards understanding. This model works great, but its limitation is in its scalability.
So, in most cases, instead of someone guiding the employee's learning and understanding, the responsibility to find the right answer to questions is laid on the employee himself. If the employee needs to learn about a new tool, technology or process in a company, he could go and search a corporate intranet, wiki or whatever sources of information the company has. It is not bad, after all, as by searching and trying to apply the information he found, the employee could learn some related things as well. Still, this way doesn't seem to be optimal as the employee needs to go through a lot of information that is not always relevant, and when he finds the answer, it is not always the most suitable answer for him if it is not linked to his previous experience.
So, what to do?, you may ask. Is there any better way to help an employee to not only find out the information, but to ensure the answer is linked to his previous experience?
In my opinion, what is needed here is a combination of sophisticated search and personalization engines.
First, a search engine will narrow the possible sources of information, like documents, web pages etc., and then a personalization engine will bring to the top those that are most relevant to the employee's previous experience.
The search engine is something that we are used to using every day, like Google. In a corporate environment, it could be a separate engine for each information system or some sort of more advanced engine, which could query many different backend systems and combine results. But how to make those results more personalized, e.g. related to the employee's previous experience? Or to be even more precise, how do we know what is the employee's previous experience?
How an employee's experience could be digitalized and made understandable for a machine?
● The first step in that direction is to list for each employee what skills he has learned previously and what is the level of knowledge. It could be done, for example, by analyzing his CV using text analytics or by asking him to fill in a simple form with skills in rows and knowledge levels in columns. This information already gives some clue, but not with much detail, though. In addition, this is a pretty much static view, which doesn't take into account the learning and experiences happening after those skills had been listed.
● The second and already better case is when, in addition to Competence Matrix, collected in the previous step, HR is collecting information about training attended, certifications achieved and everything else related to formal learning. For the employee, this could improve the relevance of the information found significantly, as the level of his knowledge is known to the system and is updated in accordance with formal learning attended. As a result, for example, expert level information will not be shown to the novice and vice versa.
Still, that is not enough to make search results truly relevant and personal. The personalization engine needs much more granular information about what the employee knows, how he prefers to learn, what is easy and what is difficult for him, what type of information suits him best etc.
How could this become digitalized and analyzed?
● The third, much more advanced option is to combine the approaches already described with collection of information about all learning activities happening in real time. What I'm talking about is the usage of Experience API (xAPI) in combination with Learning Record Store (LRS). With xAPI it is possible to collect information about the employee's learning happening in many places and with a great degree of detail.
If the company intranet is xAPI enabled, it is possible to collect information about opened pages and downloaded documents. If the employee's actions in a factory are tracked with some sensors, those could be collected as xAPI statements. If a company uses simulations or Virtual/Augmented Reality (VR/AR) for training, actions and events inside the simulation could be tracked and stored as xAPI statements. If a company uses a Learning Experience System, like Valamis, for delivery of its training, all the user's actions are automatically tracked with xAPI.
How to make learning even more personalized and effective?
When all that information about the employee is available to the personalization engine, it could make much more justified guesses about the relevance of some information for the employee. By analyzing learning history, it could become clear what learning format is most suitable for the person, e.g. does reading an article contribute more to the learning than audio, does he prefer longer learning sessions with broad context presented or should it be short and directly to the subject, do learning needs vary depending on time of day or day of week, etc.
Even more relevant results could be produced when the employee's knowledge and experience is compared to other employees' experiences and similarities found in roles, skills or learning activities. Then the relevance of the information provided to the employee could be improved based on that similarity. This also works well for a new employee who has no previous history in a company. At first, similarities like role and department as well as an absence of previous learning history could bring them relevant onboarding materials, and then by analyzing the history of the previous newcomers, a recommendation engine will be feeding the learner with suggestions that are already proven to be relevant to those who came and went through onboarding materials before.
Needless to say, it is not the end of the story, it is only just the beginning. To bring it to the next level, the loop must be closed and learning should be applied to the personalization engine to produce better relevance over time. Analyzing learning activities, checking what employees were choosing themselves from results suggested, asking were they satisfied with results provided, looking into refined searches – all that will make the recommendation engine improve and adapt all the time using machine learning.
Making learning personalized has a great impact on the learning outcomes. When new concepts are linked to a person's previous experience, it results in better understanding and learning becomes more effective.
In an organizational environment, creation of personalized learning requires technology solutions to make it cost effective and scalable. Technology, like a combination of Experience API (xAPI) and Learning Record Store (LRS), enables collection of an employee's experiences on a very granular level in digital form. This information could be then leveraged in building a personalized learning experience in future learning activities using a combination of sophisticated search and personalized engines.
Of course, none of the technology solutions is perfect, but by closing the feedback loop from a learner's activities back to the solution, the quality of answers given to a learner will be constantly improving. Having that, instead of "figure out yourself" type of learning, employees could enjoy and experience truly personalized learning.