Skip to main content icon/video/no-internet

Student modeling is the process of creating a dynamic representation of students’ personal attributes so as to provide individualized support in the learning process. The resulting profile is typically called student model or learner model. Student modeling has several parallels with user modeling, which is widely used in the commercial sector to create personal profiles of users. After discussing the use of student modeling in learning applications, this entry describes its various components and then examines both the role of student modeling in emerging learning environments and the use of learning analytics.

Use of Student Modeling in Learning Applications

In contrast to a typical learning management system that provides the same course content, activities, and interactions to all students, adaptive and personalized learning systems attempt to customize learning experiences for individual students by adapting the learning environment to their individual characteristics, competencies, interests, and needs. By doing this, these systems aim at improving learning outcomes while fostering better learning efficiencies and effectiveness. To be able to do that on an ongoing basis, these systems need to not only know as much as possible about the students but also about any changes in students’ individual attributes. Student models are, therefore, key to every adaptive and personalized learning system.

Historically, student models have been used in intelligent tutoring systems, with earlier systems focusing primarily on competency improvement for students. Later realizations of adapting learning to individual student behavior, preferences, and other characteristics evolved earlier learning systems into adaptive and personalized learning systems. Although researchers and implementers have frequently used the terms interchangeably, there are differences in the ways that adaptive learning systems and personalized learning systems play a role in improving learning experience; these differences impact the information needed in their respective student models. On the one hand, adaptive learning systems emphasize providing different course content, activities, and other aspects automatically to different students, focusing primarily on performance and progress in a course. On the other hand, personalized learning systems attempt to customize courses to suit individual student characteristics, including prior performance and progress but also interests and other relevant factors that influence learning.

Various Components of Student Modeling

Being part of a variety of learning applications, student modeling requires different types of information about the students. Major components of student modeling are described below. Depending on the type of application, certain components play a more prominent role than others.

Demography Component

The demography component includes primarily static information that can be used by the adaptive and personalized learning systems to initiate adaptive processes. Information gathered through the demography component of student modeling typically includes a student’s name, gender, student ID, begin time of study, grade, study program, and contact information.

Competency Component

The competency component of a student model contains information about a student’s domain competence. Several variations are used depending on the purpose and the situation in which the student model will be used. Four types of student models have been commonly used: overlay student model, differential student model, perturbation student model, and constraint-based student model.

Overlay Student Model

In an overlay type of student model, the knowledge of the student is assumed to be a subset of an expert’s knowledge at any point in time. The bigger the difference between student’s knowledge and expert’s knowledge, the bigger is the lack of student’s skills and knowledge. Adaptive systems target this difference to analyze what and how to provide instruction to the individual student. Different variations are possible in the overlay model, ranging from binary type where the only information recorded is whether a concept is learned or not, to a more granular approach where system records how much of the overall knowledge is learned.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading