Designing Adaptive and Personalized Learning Environments

Kinshuk

• Publisher: Routledge
• Year: 2016
• ISBN:
  9781317747703
• Edition: 1
• Topic: Education
• Subtopic: Education General

Designing Adaptive and Personalized Learning Environments provides a theoretically-based yet practical guide to systematic design processes for learning environments that provide automatic customization of learning and instruction.

The book consists of four main sections: In "Introduction and Overview, " the concepts of adaptivity and personalization are introduced and explored in detail. In "Theoretical Perspectives with Example Applications, " various theoretical concepts underlying adaptive and personalized learning are discussed, including cognitive profiling, content-based adaptivity, exploration-based adaptivity, and mobile and ubiquitous settings. In "Practical Perspectives with Example Applications, " the implementation process for adaptive and personalized learning environments is described, followed by application in various contexts. In "Validation and Future Trends, " various evaluation techniques for validating the efficiency and efficacy of adaptive and personalized learning systems are discussed. This final section concludes with a discussion of emerging trends in adaptive and personalized learning research.

Based on cutting-edge research, Designing Adaptive and Personalized Learning Environments is appropriate as a primary textbook for both undergraduate and graduate courses focused on the design of learning systems, and as a secondary textbook for a variety of courses in programs such as educational technology, instructional design, learning sciences, digital literacy, computer based systems, and STEM content fields.

part one
Introduction and Overview

one
Defining Adaptivity and Personalization

Online learning has come a long way since its emergence in mainstream education in the mid-90s. Significant growth in both hardware and software technologies in recent years have now made it a real possibility to remove barriers to education and widen the access for those who are not able to come to a physical campus. However, the majority of current implementations of learning environments lack the support individual students need to achieve success in their learning. This is not to say that there are no possibilities to support each and every student individually in achieving success in their learning. There is a whole range of tools, techniques and pedagogies that can facilitate such individual support in online learning. This book is designed to provide both conceptual understanding and practical solutions to systematically design learning environments that provide automatic customization of learning and instruction to individual learners.

Many of us would be able to recall experiences from our past, when we struggled in the classroom because the teacher’s explanations were not making sense, or we got bored because what was being taught we already knew. But it was impossible to ask the teacher to do something differently because the teacher needed to take care of the whole class and not just one student. Fast forward to today, almost every educational institution has now embraced some form of online learning, whether as a supplement to their traditional face-to-face teaching or to provide certain courses specifically in an online mode. The situation, however, has not changed.

Various types of learning environments have emerged to support different learning activities. Most popular types are content management systems (CMS), learning management systems (LMS) and social platforms. Most of these environments are primarily teacher-driven, providing various tools and functionality to support various teaching tasks. For example, teachers can create several learning units in LMS, provide various quizzes to learners, structure the whole learning sequence, and make discussion forums available to students. All students in one course then see and experience exactly the same content, navigation, presentation and other aspects of the course, despite the fact that they may be from different backgrounds, may have different preferences for learning activities, have different media preferences, already know some of the content, and have achieved different competence levels in various topics of that course. These environments, by their nature, are designed to run multiple courses at a time, involving a large number of students, and are not expected to provide a customized learning experience to individual students.

Environments with adaptivity and personalization focus on the process of learning differently. They focus on individual differences of students, and based on certain criteria, customize learning to suit individual students. Learning process in such environments does not require each and every student to follow a pre-determined rigid path. Instead, content, activities, navigation, presentation, interaction and other aspects of the course are adapted and personalized to the real-time context of the individual student.

Let us look at the two terms we have been using: adaptivity and personalization. What do they mean and how are they related to each other?

Adaptivity

When a teacher is teaching a class, adaptivity means continuously looking at the impact of teaching approaches on the students, and changing various aspects of teaching to improve student learning. In an online environment, this enables learning environments to analyze the learning processes of individual students on a continuous basis, and allows them to make modifications geared towards better learning outcomes. For example, the environment can recommend to every student who did not show sufficient competency in a quiz, to go through the associated learning unit again before proceeding to the next unit.

Personalization

Let us take an example of two students learning a particular accounting concept in an online environment. One student comes from a family running a small business. Another student comes from a family of farmers. Personalization of learning for both students would mean understanding their backgrounds and providing them with practice cases that they can relate to. For the student with the business background, a case from a factory scenario would provide a more familiar situation to relate to, whereas for the student with the farming background, a case based on a farming situation would provide better familiarity.

Dimensions of adaptivity and personalization

Adaptivity and personalization are, in some sense, two sides of the same coin. Adaptivity can be seen as a perspective from the learning environment’s side, whereas personalization considers an individual student’s perspective. Both ultimately aim to improve learning for individual students by increasing student’s efficiency, effectiveness, and satisfaction. These three aspects are not always in sync with each other. For example, increase in efficiency may require focusing only on important concepts, but that approach may affect effectiveness of the learning process and even lead to reduced satisfaction.

Compared with other scenarios, learning situations also have their unique characteristics when it comes to student efficiency, effectiveness and satisfaction. For example, in an office environment, efficiency is increased by reducing the amount of duplication. However, in learning, revision of content and practice of associated skills are important for achieving mastery.

It is important to note that the better a learning environment is able to understand the student and their learning goals, the better it will be able to provide adaptivity and personalization. At the same time, the more the student understands the capabilities and limitations of the learning environment, the more realistic his/her expectations would be towards what to expect from the learning environment. In other words, the better the student and the learning environment will understand each other, the better adaptivity and personalization will take place.

The concept of adaptivity and personalization has been around for a long time both within learning domains, and in other areas. For example, Michael Hannafin and Kyle Peck discussed characteristics of a good (effective) computer aided-instruction environment in as early as the 1980s (Hannafin & Peck, 1988). Oppermann (1994) categorized adaptive environments ranging from system initiated adaptivity with no user control, to user initiated adaptability where the environment provides various tools to the user for customizing system behaviour but does not change behaviour itself. In between these two extremes are the stages of system initiated adaptivity with pre-information to the use about the changes, user selection of adaptation from system suggested features, and user desired adaptability supported by tools and performed by the system (Oppermann, Rashev & Kinshuk, 1997).

Test your understanding

1. Which of the following would be example(s) of adaptivity and/or personalization?
   1. A teacher gives an assignment to all students in a class after three lecture classes.
   2. A teacher tells a student to study section four of chapter three again after the student received low marks in the assignment.
   3. A student searches on the Internet for suitable content after getting low marks in a test.
   4. A student retakes a course after failing it the first time.
2. Which of the following factor(s) would be suitable for consideration to provide adaptivity and/or personalization?
   1. student’s name
   2. student’s age
   3. student’s driver’s license number
   4. student’s preference for videos

Need for adaptivity and personalization

In online learning environments, it is possible for students to learn at their own pace, without the restriction of time and location. While this flexibility provides convenience to students and is seen as a positive feature, it can also create situations where students may find themselves without the presence of a teacher at the time they are learning. The online nature of learning environments also enables students from anywhere in the world to learn from the same environment. This leads to time zone differences, once again creating the possibility that an expert may not be available when students need help. Adaptivity and personalization in these scenarios aim to fill the gap created due to the absence of the teacher by providing appropriate support, help and feedback to the students. Adaptivity and personalization methods also help with customizing environment behaviour so that the students do not have to spend time learning how to use the environment instead of focusing on actual learning tasks.

How does adaptivity work?

The principle ingredient of adaptivity and personalization is that the learning environments monitor what each individual student is doing. They look for any action patterns—any sequence of actions, that can lead to problems students may face. By monitoring these action patterns, through different components of environment’s interface where the student is interacting with the environments, these environments try to find if there are ways in which, first of all, students’ errors could be corrected, and second, how students’ learning process through the environment could be improved and made more efficient. Some environments also support students in the learning phase by introducing them to various system operations. Learning environments are complex by nature. Complex environments typically have lots of different functions, and it is not easy for the users of such environments to know all the different functionalities that are available. Let us take an example of a word processing application. Not everyone is familiar with managing “reference sources” or with inserting “table of authorities”. In other words, there are so many functions in these applications that many of us may not use, even after using these applications for years. In the context of learning environments, if there was a way to find out what students really need and whether they are currently using the system functionality appropriately, the environments could provide certain help or support to the students to make them familiar with those functions. That would make students’ learning process much more efficient.

Some environments draw students’ attention to unfamiliar tools. If a student is doing an activity in a certain environment, and the adaptivity and personalization mechanism finds that there is a better way to do it because there is a readily available function or tool in the environment to do that task more efficiently, then the environment can introduce that function or tool to the student.

The main driver for adaptivity and personalization mechanism in the learning environments is the situations when students make errors during the learning process. When a student is using the learning environment appropriately, then the adaptivity and personalization mechanism remains inert. However, it keeps monitoring student’s actions. As and when the student makes a mistake in the learning process, such as selecting a wrong parameter in a simulation or incorrectly answering a quiz, the mechanism then kicks in and checks why the student made that mistake. A good adaptivity and personalization mechanism does not just inform the student that he/she has made a mistake; instead, it tries to identify why the mistake was made and what was the cause of it. It then tries to remedy that cause by recommending certain remedial content to the student, or by changing the system behaviour, such as providing content in the format more suitable for that particular student.

Benefits of adaptivity and personalization

In a typical classroom, there are lots of students and it is impossible for teachers to provide proper support to each and every student’s needs. Some students in the class would typically succeed even without the teacher. These students are well-versed in the learning process and they are typically in the top range of the class. Then there are other students who are in the bottom range and are probably not prepared for that particular course. However, there remains a large range of students in the middle who would do much better with some support, but that support needs to be customized. Students have different backgrounds, they have different preferences, and they learn differently. Individual students have different approaches regarding how they understand the material and how they can get the skills in a more effective and efficient way. Therefore, in a typical classroom with a large number of students, it is not possible for one teacher to provide all that support. That is where adaptivity and personalization comes in. The adaptivity and personalization mechanism looks at individual student’s needs, their characteristics, backgrounds, competencies, mental abilities, behavioural preferences, effective states, and so on. Based on these parameters, the adaptivity and personalization mechanism can recommend what kind of learning experience would best suit each individual student.

For example, if there is a student with a particular level of cognitive abilities, let us say, a student who has a low working memory capacity. Such a student will not be able to keep too much information at a time in his/her mind for immediate processing. On the other hand, there is another student who has a higher working memory capacity and therefore can do a lot of calculations in their mind. These students need to be taught differently. Depending on the aim of the learning process, adaptivity and personalization may help the first student who has a low working memory capacity, by providing material that does not require a lot of immediate calculations. On the other hand, the teacher may decide ‘No, this student needs to learn how to compensate for his/her low working memory capacity.’ In that case, the adaptivity and personalization mechanism may challenge that student by providing the content that requires intermediate calculations that the student has to do in his/her mind. Another example could be based on learning styles. Students have different learning styles; some students like to first get an overview of the whole material before going into detail, whereas other students like to go through the material sequentially rather than getting overwhelmed by the overview of the content. So different students have different needs because of their different characteristics, and adaptivity and personalization can provide benefit to all these different students by monitoring the needs of the individual students.

Student characteristics

Different characteristics of the students can be used in various aspects of adaptivity and personalization. For example, they can help in providing adaptive navigation guidance in the learning environment, selecting the level of granularity of the domain content to provide to the student at a certain point in time, identifying whether the student will learn effectively from analogies or not, and so on.

These student characteristics can be classified as follows:

• Behavioral attributes: These include various characteristics related to the students’ personal behaviors, such as preferences of the student, familiarity with the exploration process, and familiarity with various types of multimedia objects. Behavioral attributes can be solicited through direct inputs from the students, such as through questionnaires.
• Performance attributes: These attributes are related to the students’ competency in the subject matter, such as the level of the student’s current understanding of the domain content, his/her experiences within the domain, and competency in domain related skills. Performance attributes can be identified through quizzes, tests and other similar assessment mechanisms.
• Cognitive attributes: These attributes are related to the student’s cognitive load capacity. They require understanding of the student’s cognitive abilities. Examples include working memory capacity that allows humans to keep a limited amount of information active for a short period of time, inductive reasoning ability that enables construction of concepts from examples, information processing speed that determines how fast the students can acquire information correctly, and associative learning skill that enables linking of new knowledge to existing knowledge.
• Physiological attributes: These attributes are related to the student’s physical state. They can be analyzed by measuring various physical parameters. For example, students’ stress levels can be calculated by checking the heart rate variability. Similarly, the skin temperature, pupil dilation, rate of perspiration and other such parameters can help in understanding the student’s current physical state.

Limitations and problems of adaptivity and personalization

While adaptation has lots of benefits for students, t...

Table of contents

• Cover Page
• Half Title page
• Series page
• Title Page
• Copyright Page
• Contents
• Acknowledgements
• Part I Introduction and Overview
• Part II Theoretical Perspectives with Example Applications
• Part III Practical Perspectives with Example Applications
• Part IV Validation and Future Trends
• Index