personalized learning guide_082913
Guide to Personalized Learning Environments What IS a personalized learning environment? Race to the Top brought the phrase “personalized learning environment” into the K12 education lexicon. Before you can build (or buy) such an environment, it would be helpful to know what it is. Unfortunately the U.S. Department of Education does not define it, and there is no widely accepted definition. So, when the first district RTT competition was launched, we brainstormed with some district leaders to see if our collective experience in education and technology could paint a reasonable picture of a personalized learning environment. We hope you find this as useful as we did. Sketching a personalized learning environment We went through a lot of drawings to come up with our first personalized learning sketch:
© 2013 Amplify Education, Inc. All rights reserved.
Page 1
amplify.com
First, note the “goal” of personalized learning: student learning outcomes. The drawing includes prescribed outcomes by the school (district, state), as well as selected outcomes by the student. In a personalized learning environment, students have choice, but choices about what they choose to master augment rather than replace the college- and career-ready mastery objectives of the LEA. Given the connection to RTT, we started with the RTT four assurances, and the district’s own strategic vision, as a “given.” But for a base, we decided that a rich student model was the critical foundational element. A rich student model is a data model of each student, which includes information related to their academic history, demographic bio, learning preferences and interests. It needs to not only take in data from disparate systems (no small matter, that), but also be flexible enough to encompass data from systems such as educational games where student activity data could provide critical insights. Key to personalizing learning paths are algorithms. We don’t need to define algorithm, but we do need to recognize that we are a long way from having computer algorithms that have been proven in the classroom, at scale, to recommend appropriate learning paths for a variety of students in a variety of situations. As a result,
At Amplify, we also believe that algorithms need to be partly based on a theory-of-learning rather than blindly empirical, such as you find with web-based
flexibility in the personalized learning environment to
analytics to drive consumer
evolve the algorithms — as we have more data connecting
spending.
student outcomes with inputs such as curriculum and interests — is important. By wiring learning experiences so that they capture as much data as possible, you can figure out how to optimize learning paths along numerous dimensions, such as difficulty, interests, confidence levels, persistence and learning styles. Systems with rich student models and appropriate algorithms can respond to the reality of a given student, a given teacher, a given school at a given moment – rather than some average of these parameters. One often hears that the point of a personalized learning system is to maximize the time that students spend in their “zone of proximal development” (ZPD). But it is important for students on occasion to be stretched above the zone, as well as to feel masterful below the zone. Further, the social value of a group working together may outweigh the importance of each student being in their own ZPD, allowing, for example, weaker students to grow faster than they would on their own. Finally, it shouldn’t always be “the system” that places students in their ZPD. Often the teacher should do it (informed by data as well as their own observations). Master teachers, of course,
© 2013 Amplify Education, Inc. All rights reserved.
Page 2
have developed their own implicit “algorithms” for providing individualized instruction to their students. Teachers, and groups of teachers, are the next critical layer in this ecosystem. The environment needs to take into account the observations and judgments of these teachers, as well as provide critical support such as real-time, in-place professional development (PD) and classroom management tools that support teacher-student interaction, rather than replace such interaction. With this structure of systems, data, and staff, only one thing is missing: what we call an ecosystem of great teaching and learning stuff. This “stuff” includes the kinds of normal content and curriculum found in classrooms and what students today access on the Internet. But we all know that “stuff” includes the good, the bad and the downright ugly. That’s why the key word in this part of the personalized learning environment is “great.” Discovering great stuff for teaching and learning is hard work, and creating it is even harder. When we look at textbooks, whether of the e- or p-variety, we don’t see great stuff — we see rehashed and repurposed content, we see watered-down and dumbed-down content. Students, or groups of students with similar proclivities or capabilities, can now have their personalized learning pathway through this great stuff. Built on the four assurances and the layers described, these pathways will lead to greater learning outcomes for students of all interests and capabilities. What about those Onions? The second personalized learning sketch we call “Onions.” This is a different way of looking at the learning paths of students enabled by the personalized learning environment. If you assume that most students in most schools in most districts will still be in a traditional classroom setting along with 25–30 students (and likely more over time) and one teacher (at least for the near-term) then what does a personalized learning path look like? We envisioned a series of onion-like units — each with different layers that represent different “paths” through the same content, some less challenging, some more — but all leading toward the same learning objectives (and hence the same learning assessments). Of course, small groups of students may go outside the “onion” entirely — as if they were in a home-school setting — but still aim at the same endpoint.
© 2013 Amplify Education, Inc. All rights reserved.
Page 3
Feedback Advancements in neuroscience have shown us the amazing plasticity of the human brain, with insights into how neural pathways are created and strengthened in the learning process. Recent research has shown that “feedback has the potential to have one of the most powerful influences on student learning, with an effect size twice that of all in-school effects.”1 Of course, this is both common sense and something all really good teachers already know. But if you observe a lot of classrooms, as we have, you will see how little individual feedback is actually given in a typical 50-minute period. And that feedback is often parceled out to a small handful of students.
Look what just landed… Giving quality, timely in-classroom feedback is hard, and that is why we’ve built Hummingbird, a feedback management tool for teachers to use on an iPad or a Access Tablet that will revolutionize how they manage in-class writing activities and sharing activities, not only improving students’ results but teachers’ teaching as well!
So, a personalized learning environment can bring substantial teaching and learning value if it can facilitate more and better feedback by teachers, peers, and even by intelligent learning and formative assessment systems targeted at students when and where they need it. The diagram 1
Hattie, J. (2012).Visible learning for teachers: Maximizing impact on learning. Routledge.
© 2013 Amplify Education, Inc. All rights reserved.
Page 4
below is one view of a how this might look, with different sources of feedback as a key driver for students’ personalized learning paths.
You’ll note on each of these diagrams the commentary regarding continuous improvement, which goes without saying is important but of course is difficult to design. Even more of a hand-wave is the comment “new arrangements of time, people, money, and facilities optimized for achievement, engagement, and effort.” These fall outside the scope of the personalized learning environment, but not outside the scope of anyone serious about creating a 21st century learning environment. To take a simple example, the opportunities for quality feedback from teachers to students, and students to students, could increase substantially if class periods were rearranged for longer blocks, as many schools are already doing. Conclusions A school’s personalization strategy needs to be optimized for many different ends – teacherstudent ratio, durability, reliability, scale, fun, social dynamics, time, interdisciplinary value, local community needs and more. Taken together, all this may seem daunting — and going from handdrawn illustrations to reality represents a lot of work by a lot of people over many years. But, it might help to focus on a simple way of judging a learning experience: the extent to which it inspires students to do more work. This is not just because work is good but because willingness to do work is often an indication of engagement and many other things going right.
© 2013 Amplify Education, Inc. All rights reserved.
Page 5
A final word from our sponsor:
“Results only happen if we significantly alter what teachers and students work on, dramatically increase their motivation to do the work, radically intensify the attention they apply to it, and massively increase the time they spend doing it” – this is Amplify’s focus.
© 2013 Amplify Education, Inc. All rights reserved.
Page 6
© 2013 Amplify Education, Inc. All rights reserved.
Page 1
amplify.com
First, note the “goal” of personalized learning: student learning outcomes. The drawing includes prescribed outcomes by the school (district, state), as well as selected outcomes by the student. In a personalized learning environment, students have choice, but choices about what they choose to master augment rather than replace the college- and career-ready mastery objectives of the LEA. Given the connection to RTT, we started with the RTT four assurances, and the district’s own strategic vision, as a “given.” But for a base, we decided that a rich student model was the critical foundational element. A rich student model is a data model of each student, which includes information related to their academic history, demographic bio, learning preferences and interests. It needs to not only take in data from disparate systems (no small matter, that), but also be flexible enough to encompass data from systems such as educational games where student activity data could provide critical insights. Key to personalizing learning paths are algorithms. We don’t need to define algorithm, but we do need to recognize that we are a long way from having computer algorithms that have been proven in the classroom, at scale, to recommend appropriate learning paths for a variety of students in a variety of situations. As a result,
At Amplify, we also believe that algorithms need to be partly based on a theory-of-learning rather than blindly empirical, such as you find with web-based
flexibility in the personalized learning environment to
analytics to drive consumer
evolve the algorithms — as we have more data connecting
spending.
student outcomes with inputs such as curriculum and interests — is important. By wiring learning experiences so that they capture as much data as possible, you can figure out how to optimize learning paths along numerous dimensions, such as difficulty, interests, confidence levels, persistence and learning styles. Systems with rich student models and appropriate algorithms can respond to the reality of a given student, a given teacher, a given school at a given moment – rather than some average of these parameters. One often hears that the point of a personalized learning system is to maximize the time that students spend in their “zone of proximal development” (ZPD). But it is important for students on occasion to be stretched above the zone, as well as to feel masterful below the zone. Further, the social value of a group working together may outweigh the importance of each student being in their own ZPD, allowing, for example, weaker students to grow faster than they would on their own. Finally, it shouldn’t always be “the system” that places students in their ZPD. Often the teacher should do it (informed by data as well as their own observations). Master teachers, of course,
© 2013 Amplify Education, Inc. All rights reserved.
Page 2
have developed their own implicit “algorithms” for providing individualized instruction to their students. Teachers, and groups of teachers, are the next critical layer in this ecosystem. The environment needs to take into account the observations and judgments of these teachers, as well as provide critical support such as real-time, in-place professional development (PD) and classroom management tools that support teacher-student interaction, rather than replace such interaction. With this structure of systems, data, and staff, only one thing is missing: what we call an ecosystem of great teaching and learning stuff. This “stuff” includes the kinds of normal content and curriculum found in classrooms and what students today access on the Internet. But we all know that “stuff” includes the good, the bad and the downright ugly. That’s why the key word in this part of the personalized learning environment is “great.” Discovering great stuff for teaching and learning is hard work, and creating it is even harder. When we look at textbooks, whether of the e- or p-variety, we don’t see great stuff — we see rehashed and repurposed content, we see watered-down and dumbed-down content. Students, or groups of students with similar proclivities or capabilities, can now have their personalized learning pathway through this great stuff. Built on the four assurances and the layers described, these pathways will lead to greater learning outcomes for students of all interests and capabilities. What about those Onions? The second personalized learning sketch we call “Onions.” This is a different way of looking at the learning paths of students enabled by the personalized learning environment. If you assume that most students in most schools in most districts will still be in a traditional classroom setting along with 25–30 students (and likely more over time) and one teacher (at least for the near-term) then what does a personalized learning path look like? We envisioned a series of onion-like units — each with different layers that represent different “paths” through the same content, some less challenging, some more — but all leading toward the same learning objectives (and hence the same learning assessments). Of course, small groups of students may go outside the “onion” entirely — as if they were in a home-school setting — but still aim at the same endpoint.
© 2013 Amplify Education, Inc. All rights reserved.
Page 3
Feedback Advancements in neuroscience have shown us the amazing plasticity of the human brain, with insights into how neural pathways are created and strengthened in the learning process. Recent research has shown that “feedback has the potential to have one of the most powerful influences on student learning, with an effect size twice that of all in-school effects.”1 Of course, this is both common sense and something all really good teachers already know. But if you observe a lot of classrooms, as we have, you will see how little individual feedback is actually given in a typical 50-minute period. And that feedback is often parceled out to a small handful of students.
Look what just landed… Giving quality, timely in-classroom feedback is hard, and that is why we’ve built Hummingbird, a feedback management tool for teachers to use on an iPad or a Access Tablet that will revolutionize how they manage in-class writing activities and sharing activities, not only improving students’ results but teachers’ teaching as well!
So, a personalized learning environment can bring substantial teaching and learning value if it can facilitate more and better feedback by teachers, peers, and even by intelligent learning and formative assessment systems targeted at students when and where they need it. The diagram 1
Hattie, J. (2012).Visible learning for teachers: Maximizing impact on learning. Routledge.
© 2013 Amplify Education, Inc. All rights reserved.
Page 4
below is one view of a how this might look, with different sources of feedback as a key driver for students’ personalized learning paths.
You’ll note on each of these diagrams the commentary regarding continuous improvement, which goes without saying is important but of course is difficult to design. Even more of a hand-wave is the comment “new arrangements of time, people, money, and facilities optimized for achievement, engagement, and effort.” These fall outside the scope of the personalized learning environment, but not outside the scope of anyone serious about creating a 21st century learning environment. To take a simple example, the opportunities for quality feedback from teachers to students, and students to students, could increase substantially if class periods were rearranged for longer blocks, as many schools are already doing. Conclusions A school’s personalization strategy needs to be optimized for many different ends – teacherstudent ratio, durability, reliability, scale, fun, social dynamics, time, interdisciplinary value, local community needs and more. Taken together, all this may seem daunting — and going from handdrawn illustrations to reality represents a lot of work by a lot of people over many years. But, it might help to focus on a simple way of judging a learning experience: the extent to which it inspires students to do more work. This is not just because work is good but because willingness to do work is often an indication of engagement and many other things going right.
© 2013 Amplify Education, Inc. All rights reserved.
Page 5
A final word from our sponsor:
“Results only happen if we significantly alter what teachers and students work on, dramatically increase their motivation to do the work, radically intensify the attention they apply to it, and massively increase the time they spend doing it” – this is Amplify’s focus.
© 2013 Amplify Education, Inc. All rights reserved.
Page 6
