Personalised learning – what the hell is it? 10 things that work...

In one sense all learning is personal as it requires personalattention, effort and practice. But ‘Personalised Learning’ as a term has come to mean the tailoring of learning to the individual, with a sensitivity to their individual needs. The goal being to increase the efficacy of teaching and learning, lower failure rates and increase access with scalable persoanlisation. 

Myths

First, we need to dispense with the myth that this is about Learning Styles or some other phantom phenomenon. This is one of the commonest myths in education, deeply embedded but completely wrong-headed. Personal needs are far best defined in terms of a whole raft of data about what you know, need to know and your personal context and circumstances. I like to think of Personal learning as a solution to learning difficulties. We all have learning difficulties, as our brains are inattentive, get easily overloaded, forget most of what we’re taught, sleep 8 hours a day, are hopelessly biased, can’t download and can’t network. Then there’s the delivery problems associated with the current model, which is expensive, time-consuming, inefficient and non-scalable. If, as some believe, personalised learning offers a way forward for technology enhanced learning, then let’s give it a chance.

Bloom’s promise

Personalised learning can be delivered offline, but it requires one-to-one tuition. Hence the great interest in Bloom’s famous paper, The 2 Sigma Problem, which compared the lecture, formative feedback lecture and one-to-one tuition. Taking the straight lecture as the mean, he found an:

84% increase in mastery above the mean for a formative approach to teaching and an astonishing 

98% increase in mastery for one-to-one tuition. 

Google’s Peter Norvig famously said that if you only have to read one paper to support online learning, this is it. In other words, the increase in efficacy for one-to-one, because of the increase in relevant and on-task learning, is immense. This paper deserves to be read by anyone looking at improving the efficacy of learning, as it shows hugely significant improvements by simply altering the way teachers interact with learners. 

Personalised promise through technology

However, given the difficulty in scaling up through human one-to-one tutoring, the promise of Personalised Learning is now seen as being practically realised through technology. There’s a whole range of technologies that have emerged over the last few years, that have accelerated progress right across the learning journey – interface advances, engagement, support, content, curation, practice, assessment, tutoring, even wellbeing. Rather than focus on presentation, AI tends to rely more on smart software, to deliver more than just souped-up graphics and gamification. It relies on smart software, AI, to deliver personalised learning, not as one thing but as many things in the complex business of learning.

1. Personal interface

First up is a general move in interface technology, where AI is the new UI. Google, Facebook, Twitter, Amazon, Netflix and almost everything we do online is mediated by AI. Interaction is shifting towards ‘voice’, which is AI-driven. This, we can expect, to have some influence on the delivery of learning. It is even possible that AI-driven speech will radically change learning delivery. At WildFire, we are using voice only navigation and input by learners for the entire learning experience. We create podcasts on the fly, using AI. Alexa, Google Home and other consumer devices use voice as both input and output. Interfaces are becoming more frictionless, which reduced cognitive load for learners making learning more frictionless.

2. Personal engagement

Learners are lazy. There I’ve said it. Procrastination is the norm. Most do not do the pre-reading, leave assignments and essays until the last minute, miss lectures and behaviour in schools can be a real problem. Bots, such as Differ, can be used to engage and nudge students. This ‘push’ side of tech, whether by bots, email or message apps, can be personalised to match the individual’s pattern of needs and behaviour. Face recognition tech can also be used for registration, even spotting learning behaviours. In general, we know how engaging tech is for learners – as we spend most of our time complaining about its distractive effects. Suppose we take some of those engagement tricks and apply them to learning.

3. Personal support

Direct support though bots have already had success as Georgia Tech, where the students not only mistook the bot for an academic teacher but put it up for a teaching award. Bots, such as Otto, embedded on the workflow do a similar job in corporate learning. Support in online learning may come through access to resources, content and people but if it knows what your personal needs are, it helps. 

4. Personal content (adaptive)

The personalised delivery of appropriate learning experiences, based on what you need at that exact time, drawn from data on you, others taking the course, even your (and others) behaviour on previous courses, can all be used to decide what you need at that exact moment. This ‘adaptive’ learning is almost certain to increase the efficacy of online learning. It resequences learning, and with formative feedback, can deliver on the promise of personalised learning, educating everyone uniquely. Having been involved in this area, for a number of years, I’ve seen some convincing results from real students in real institutions. This hold great promise.

5. Personal text analysis

It is unfortunate that in most institutions, the only place you’ll find AI technology, in a personalised sense, is in plagiarism checkers, seeing if you have cheated. Actually, what many are waking up to is the possibility of this same technology being used to help learners complete essays and assignments through formative feedback, using a repeated submit and learn model. Personal feedback on specific aspects of your performance, whether writing or performing in a domain specific assignment, could be one of the big wins. Formative feedback accelerates learning.

6. Personal curation

Learning experiences, especially online, often feel over-directed and fixed, a little straightjacketed. Imagine a system that uses AI to automatically provide links to useful outside sources relevant to you at that very moment you need them, as it has spotted that you have a problem. We do this in WildFire, where links to outside resources are created on the fly, making the course more porous and to encourage curiosity.

WildFire

7. Personal practice

An often forgotten dimension in personalised learning is personalised practice. Spaced practice can be delivered in a logarithmic fashion, based precisely on your needs. Learn now and between now and your exam, new job, whatever, you get a personalised pattern of spaced practice. We have delivered this in WildFire, where you get several bites of the cherry after completing the course, delivered to you by email, to shunt learning from working to long-term memory.

8. Personal assessment

Both personal formative and summative assessment are benefitting from AI. Digital identification, through keyboard strokes (Coursera), face recognition and so on, makes sure that the right person is sitting the exam. Formative assessment, often using AI techniques, is being delivered to large numbers of students. Adaptive testing is also available, to provide personalised assessments. We’ve used assessment bots in WildFire to question learners after their more directed learning experiences. Assessment is by definition personal, it’s about time we made it more personal through technology.

9. Personalised tutors

AI is nowhere near being able to replace teachers. But as the technology begins to know you, talk to you, remember what you said and did and understand the context; then we can expect services like Alexa and tutors to bloom in learning. We see this with Duolingo and many other online services online. They get to know you and deliver what you want at that very moment, free from the tyranny of time and space – anyone, anytime, anyplace. This is what Amazon does for shopping, it’s what Cleo and other bots do in finance, it’s what fitness bots do in health. We can expect moves in this direction in learning. Note that this has nothing to do with robot teachers. That idea is as stupid as having a robot driving a self-driving car.

10. Personal difficulties

We have evidence that young people are feeling the pressure and the statistics for stress, mental health problems, even suicide are rising. We also know that young learners hide this from their parents, teachers and tutors. One solution is to anonymise help. Woebot and Elli have already been used and subjected to clinical trials. The results are encouraging. Beyond this, we already know that personalised learning is a boon to learners with learning difficulties. Individualised responses to physical and psychological issues is now normal in education.

Conclusion

Personalised learning is not one thing, it is many things, an umbrella term for all sorts of applications, especially those driven by AI. We are on the cusp of bringing smart software to deliver smart learning to make people smarter. It will take time, as AI is, at present, only good at very specific, narrow tasks, not general teaching skills. That’s why I’ve tried to break it down into to specific sweet spots. But on one thing I’m clear – this is a worthy path to follow.

Bibliography

Bloom, B. (1984). The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring, Educational Researcher, 13:6(4-16).

Bloom (1913-1999) - Mastery learning. Taxonomy of learning… not a hierarchy…

Benjamin Bloom’s is best known for his ‘taxonomy’ of learning, in Taxonomy of educational objectives: The classification of educational goals: Handbook I, cognitive domain (1956)but it has not stood the test of time as well as other aspects of his work around time to competence and personalised learning. Bloom claimed that his ‘taxonomy’ book was one of the most cited and least read books in education. The taxonomy was created to validate assessment items and learning objectives. It was designed to be used as an analytic tool to create balance in courses, balanced objectives and assessments. It was a regulating tool. Instead it was turned into a partial and misleading pyramid.

Bloom’s taxonomy

Bloom is known globally for his hugely influential classification of learning behaviours and provided concrete measures for identifying different levels of learning. In the 1950s there was a movement to identify a taxonomy of learning, as examinations were being used nationally in education. Bloom chaired the committee and they saw the top level of a taxonomy as three overlapping domains:

Cognitive domain (knowledge-based)

In the original version of the taxonomy, the cognitive domain is broken into the following six levels of objectives; Knowledge, Comprehension, Application, Analysis, Synthesis and Application.

Psychomotor domain (action-based)

Skills in the psychomotor domain describe the ability to physically manipulate a tool or instrument like a hand or a hammer. Psychomotor objectives usually focus on change and/or development in behavior and/or skills.

Affective domain (emotion-based)

Skills in the affective domain describe the way people react emotionally and their ability to feel other living things' pain or joy. Affective objectives typically target the awareness and growth in attitudes, emotion, and feelings.

Taxonomy gone wrong

The taxonomy was devised to assist teachers to classify educational goals and plan and evaluate learning experiences. Unfortunately, he only published the first book on the Cognitive domain, co-authored with David Krathwohl, in 1956. It was another eight years until the second book on the Affective Domain was published. This led to an unnatural, over-emphasis on the cognitive side, captured meme-like by a coloured pyramid, that was never in Bloom’s work. His pupil, Anderson (2001) tried to correct these misconceptions, turning the nouns into verbs but the mythical Bloom endures.

Critique

In the 2001 revised edition of Bloom's taxonomy, written by Lori Anderson decades after the original publication (he had never read it), updated the taxonomy based on a more sophisticated understanding of cognitive psychology. The levels are slightly different and he abandoned the pyramid for a rectangle:

 

• Remember

• Understand

• Apply

• Analyze

• Evaluate

• Create (rather than Synthesise)

Note that there is no sense of higher and lower here, as they are often represented in diagrams. There is no hierarchy here. Unfortunately, they are still both misrepresented by coloured pyramids, that are over-hierarchical and devalue and diminish the role of knowledge in learning, placed at the bottom of the pile. Learning is neither a hierarchy of separate entities nor a linear process. It is a gross oversimplification.

It has been argued that Bloom’s taxonomy has done a lot of damage to teaching and learning. Learners from poorer backgrounds may have suffered badly from this devaluation of knowledge, as it is a necessary condition and the foundation for learning. That is not to say that all learning is remembering. Analysis and generated work are also important but this is not a linear process of climbing a pyramid, it is about an integrated approach. In reality, learning is non-linear and messier than many taxonomies and instructional theories suggest. It uses many of these activities in a more elaborate and networked fashion. The problem may very well be, say teaching the grammar and language of a language or coding before ever getting learners to actually practice dialogue or actually synthesize and code with a goal.

Indeed, we have had dozens of taxonomies that sliced and diced learning in all sorts of ways. These include; L. Dee Fink’s Taxonomy of Significant Learning Outcomes that goes beyond cognitive processes and includes other aims of teaching: Foundational Knowledge, Application, Integration, Human Dimension, Caring and Learning How to Learn; Wiggins and McTighe backwards design model describes Six Facets of Understanding: Explain, Interpret, Apply, Have perspective, Empathize and Have self-knowledge; Davis and Arend provide yet another categorization that can help educators determine which teaching methods are best suited for which learning objectives: Acquiring knowledge, Building skills, Developing critical, creative, dialogical thinking, Cultivating problem solving and decision-making abilities, Exploring attitudes, feelings and perspectives, Practicing professional judgment and Self-discovery and personal growth.

But Bloom’s, and other taxonomies, don’t pick up on contemporary findings from cognitive science, such as retrieval and spaced practice or technical skills.

Mastery learning

Bloom’s other research led him to believe, in Human Characteristics and School Learning (1976), that learners could master knowledge and skills given enough time. It is not that learners are good or bad but fast and slow. The artificial constraint of time in timed periods of learning, timetables and fixed point final exams, as a destructive filter on most. The solution was to loosen up on time to democratise learning to suit the many not the few. Learning is a process not a timed event. Learning, free from the tyranny of time. allows learners to proceed at their own pace.

Bloom proposed three things could make mastery learning fulfil its potential:

1. Entry diagnosis and adaption (50%) - diagnose, predict and recommend

2. Address maturation (25%) - personalise and adapt

3. Instructional methods (25%) - match to different types of learning experiences and time

Personalised learning

His true legacy, which is now starting to be realised and implemented through the use of technology, is personalised learning. Google’s Peter Norvig famously said that if you only have to read one paper to support online learning, this is it - The 2 Sigma Problem, where he compared the lecture, formative feedback lecture and one-to-one tuition. Taking the straight lecture as the mean, he found an 84% increase in mastery above the mean for a formative approach to teaching and an astonishing 98% increase in mastery for one-to-one tuition. In other words, the increase in efficacy for one-to-one learning, because of the increase in on-task learning, is immense. This paper deserves to be read by anyone looking at improving the efficacy of learning as it shows hugely significant improvements by simply altering the way teachers interact with learners. Online learning, in the widest sense of the word, promises what Bloom called ‘one-to-one learning’, whether it’s through self-paced structured learning, scenario-based learning, simulations or increasingly adaptive and personalised learning. The advances and availability of AI has made this possible, as explained in AI for learning, Clark (2020).

Influence

Bloom was also the first to really establish a solid, working taxonomy of learning. Unfortunately it was truncated and excluded the psychomotor and affective components, was caricatured as a hierarchical pyramid, its subsequent improvements ignored, leading to an overemphasis on the cognitive and artificially hierarchical and sequential nature of learning. It was also used to reinforce the prejudices among the academically minded, those who see knowledge, the affective and psychomotor and vocational learning as playing a diminished role in funded learning. One could argue that it has also led to an over-emphasis on 21st C skills, which explicitly devalue the role of knowledge in learning.

His longer legacy is more likely to be, not his outdated, partial and misleading taxonomy but how work on personalised learning. Free from the tyranny of time, he thinks mastery learning will produce better results for the majority of learners.

This is now manifesting itself in personalised learning technology that uses aggregated and personal data to deliver the right thing to the right learners at the right time. Resequencing content, based on continuous assessment of their personal progress is the realisation of Bloom’s research and dream of a massively more efficient and optimised learning process.

Bibliography

Bloom, B. (1984). The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring, Educational Researcher, 13:6(4-16).

Bloom, B.S. (Ed.) (1956). Taxonomy of educational objectives: The classification of educational goals: Handbook I, cognitive domain. Longmans, Green.

Guskey, T. R. (2005).Benjamin S. Bloom: Portraits of an educator. Lanham, MD: Rowman & Littlefield Education.

7 ways AI & Data are transforming learning

Tesla passed $1 trillion market cap today so it is now worth more than Pfizer, Aztrazeneca, GSK, ExxonMobil, BP, and IBM combined. The only companies now worth more than Tesla are Apple, Microsoft, Google and Amazon. Their common denominator is that their underlying tech is now AI. Europe is falling behind, as we'd rather regulate than innovate.

Those who claim to ‘know’ where AI is going, and how fast, are being constantly challenged. 

So where is it going on learning? Well the main area of focus is NLP (Natural Language processing). AI is moving fast on several fronts here.

Data

Tesla has what seems to be an outrageous valuation. Yet what is being valued is not traditional car production, it is the driving data it harvests and the promise of a world where the very concept of vehicles and transport will be transformed. This will happen in learning. The data we gather will feed into optimising future learning experiences, as processes not events. This is why AI, or rather AI that uses data, will shape the future landscape of learning. Data will lie at the heart of all learning experiences. I explain this in my new book ‘Learning Experience Design’.

AI is the new UI

I’ve written about this in ‘AI for Learning’ and ‘Learning Experience Design’, the reshaping of UX as almost all interfaces are now mediated by AI - all social media, Netflix, Amazon, Google, YouTube - almost everything you do online. This is now happening in learning thorough LXP systems. In addition, voice interfaces are now in smartphones and on devices in cars and homes. It is getting better, faster and is scalable. AI is changing our whole relationship with technology, making it more human.

AI personalises

We know that personalised learning gives really significant advantages to large numbers of learners. We’d all love to have one-on-one teaching but that was never economically possible. It is now. Adaptive and personalised learning, enabled by AI, is now here at all levels in learning. CogBooks, a company I helped build has just been sold to Cambridge University Online and will power its online learning. LXPs, such as Learning Pool’s Stream, something I’ve been involved in, will deliver personalised learning to employees in the workplace and workflow.

AI teaches

Teaching largely addresses deficits in motivation and effort, learning is largely achieved by oneself. Took me a long time to truly understand this. It can create sense-making experiences for learners. The problem with traditional online learning is that it was essentially the presentation of content. It never really did what a good teacher does and that is create the opportunities for learning then allow and support you to make the effort to learn. AI enables both. We do this in WildFire.

AI learns

We used to have teachers and learners. Now we have teachers, learners and technology that also learns. Tesla learns as it aggregates driving data and uses that data to improve performance. The more we use Google the better it gets. The more we use personalised and adaptive learning the better it becomes for future students. We are no longer stuck on a plateau of human performance but on an upward trajectory of performance, making learning better, faster and cheaper.

Transformers

Transformers, such as GTP-3 are already useful in learning. We’ve been using them in WildFire for summarisation, content creation and question generation. This software is so powerful that just learning how to ask it questions or do things for you needs a new skillset - it is called ‘prompting’. These AI models have been trained with unimaginably large data sets. They have so much data in their training set that they, at times, transcend the ability of humans to create prose. They are now also entering the world of audio, images and video. They will literally be transformative.

Edge AI

The processing and application of AI on the ‘edge’, on devices, has really arrived. Look at the new Pixel6 mobile phone to see how AI is being delivered via chips in devices such as phones. It has a Tensor AI chip on-board; so translates, transcribes and does speech recognition blazingly fast. It can also erase unwanted objects on photos. These are seriously difficult tasks that require localised processing.

Conclusion

We can wallow in existing practices and technology and see modest but not substantial change in the efficacy and cost of learning. Or we can accept that the future is one where data, and what we do with that data, determines upward progress. A future that uses AI and data to create learning experiences as processes not events, improve interfaces, personalise, teach, support learning. All of this possible to wherever, whenever and to whoever. Technology, specifically AI and Data are finally delivering what we used to call Lifelong Learning.

Solid paper on personalised learning

A solid paper on the advantages for of self-paced learning. The promise of AI is to deliver personalised learning, sensitive to the learner’s AR (acquisition rate). 

The acquisition rate in learning refers to the speed or efficiency with which a learner acquires new knowledge, skills, or behaviors. It is a measure of how quickly a person can grasp new information or master new concepts. This rate can vary significantly depending on factors such as the complexity of the material, the learner’s prior knowledge, motivation, cognitive abilities, and the methods of instruction being used.

STUDY

tested different ways of teaching multiplication tables (6s, 7s, and 8s) to 3rd & 4th graders

split into three groups: 

• Group 1: learned 2 facts at a time

• Group 2: learned 8 facts at a time

• Group 3: learned facts based on their personal "acquisition rate" (how much they could handle)

RESULTS

Key Findings (from the table and text):

Learning time varied a lot: 

• 2 facts group: 3 minutes

• Personal pace group: 7 minutes

• 8 facts group: 14 minutes

What worked best: 

• Personal pace method (acquisition rate) worked really well

• Teaching 8 facts at once was least effective (only 39% remembered)

• Teaching 2 facts at a time was middle ground (47% remembered)

• Students remembered 76% of facts when taught at their personal pace

One size does not fit all when it comes to learning multiplication facts. The average student could handle learning about 4 new facts at a time. Tailoring the number of facts to each student's learning capacity was most effective. Personalised, self-paced learning works.

Bloom's lesser known research

Bloom is known for his pyramid (he never drew one and it is hopelessly simplistic) but researched this in detail. His other research, rarely known, led him to believe, in Human Characteristics and School Learning (1976), that learners could master knowledge and skills given enough time. It is not that learners are good or bad but fast and slow. The artificial constraint of time in timed periods of learning, timetables and fixed-point final exams, is a destructive filter on many. The solution is to loosen up on time to democratise learning to suit the many not the few. Learning is a process not a timed event. Learning, free from the tyranny of time, allows learners to proceed at their own pace.

Bloom proposed three things could make mastery learning fulfil its potential:

1. Entry diagnosis and adaption (50%) - diagnose, predict and recommend. David Asubel made the point that “The most important single factor influencing learning is what the learner already knows” yet pre-testing is relatively rare form of assessment.

2. Address maturation (25%) - personalise and adapt. AI can deliver personalised learning based on continuous assessment and adaption.

3. Instructional methods (25%) - match to different types of learning experiences and time. Sensitivity to the type of learning, a far more complex issue that Blook thought with his falsely progressive pyramid and tripartite distinction of Cognitive domain (knowledge-based), Psychomotor (action-based) and Affective (emotion-based)

https://www.researchgate.net/figure/Retention-and-Efficiency-Data-by-Condition_tbl1_281801350

How to prompt like a PRO! 100 types of prompt in ChatGPT for learning

We have been stuck in a productivity rut for some time. AI promises to release us from ploughing that particular furrow, and release technology driven learning from its current malaise. AI is clearly the technology of the age. Almost all of the large, global tech companies are, in essence, AI companies.

In learning AI changes life and work, and by consequence, what, whenever, when and how we learn. AI promises to deliver adaptive and personalised learning on any topic at low cost freeing education from expensive scarcity. This is because it increases the efficacy, not only of content creation but, more importantly delivery. We see a glimpse of this with Duolingo, now an AI company with a highly effective personalised learning system, that is free. They have now expanded into Mathematics. 

They used to say that information wants to be free. A far better mantra is to say that education wants to be free. First, free from the tyranny of time, having fixed lengths of time to get to competence, fixed timetables, course times and so on. Secondly, free learners from the tyranny of place, having to be somewhere specific, like a lecture hall or classroom.

I explored this in my book ‘AI for Learning’. Accelerating learning, making it readily available, along with semantic search and the many other gifts that AI will bring, should make organisational development that much easier, cheaper, faster and better. This has been the promise of most technology and if we look at the past, technology has, on the whole, delivered. 

Generative AI

Generative AI is still in its infancy, yet achieved instant global awareness with ChatGPT. That impact came from the realisation, often from the first output to the first question you asked, that this has huge potential. It put AI into the hands of millions, made it real. There was the expected negativity, largely born of fear, as is usually the case with new technology in learning, but that quickly gave way to wonder, lots of creative uses and speculation about the future.

It is important not to see generative AI as the only way AI has and will impact L&D. My books ‘AI for Learning’ , 'Learning Experience Design' and ‘Learning technologies’ go across the entire learning journey, showing how AI has already embedded itself into almost everything we do online. In that sense it has been here for at least 20 years in search, in UI, text to speech, speech to text (increasing accessibility), also in learning support, content creation, recommended learning pathways, adaptive, personalised learning, assessment and spaced practice.

Prompt Engineering

What ChatGPT does is allow you to use the tool to teach and learn. Once you have mastered ‘prompt engineering’, knowing how to construct the right input, understanding that ChatGPT is a dialogue system not just a producer of monolithic pieces of text, you can output wonderful things. This is a new skill for learning designers and generative AI allows all sorts of complex prompting using iterations, logic and parameters that improve the output. 

It allows you to improve up-front design, get the juices going with stakeholders, producing objectives, likely competencies and skills, syllabi, even titles for your learning initiatives. To create content one can prompt for full content with the right level of detail and nuance, summaries, images, all in different styles and voices, suitable for different audiences. You can create assessments and assignments with full rubrics for marking, also learning activities for discussions, scenarios and role playing. Beyond this you can prompt for emails, social media content and marketing. We have put together a list of 100 prompt ideas for learning professionals  to allow learning professionals to widen their perspective on this technology which we will be using in planned talks and workshops.

Remember that this is only the start of AI tools that will dramatically improve productivity for teachers, learning designers and learners.