Pedagogical Module

 < Day Day Up > 



The pedagogical module consists of a set of specifications of what instructional materials the system should present and how and when it should present the materials. The pedagogical module should behave like an expert teacher who provides timely feedback, examples, analogies, multiple views, and different levels of explanations (Clancey, 1987). The system must monitor the student’s actions, advise the student about errors, and anticipate future actions based on inferences about the student’s current activities (Woolf, 1987). The system should know how to ask the right questions and how to focus on the appropriate issues. In intelligent tutoring systems, two methods are usually used when interacting with students: the Socratic and the coaching methods. In the Socratic method, students are provided with questions that guide them through the process of debugging their misconceptions. In the debugging process, students are assumed to reason about what they know and do not know and thereby to modify their conceptions. The coaching method provides students with an environment in which to engage in activities such as computer games and simulations in order to learn related skills and general problem-solving skills. The coaching method is developed to identify diagnostic strategies required to infer a student’s misunderstanding from observed behaviors and to identify various explicit tutoring strategies for directing the tutor to say the right thing at the right time.

There are many learning theories on how humans learn and process information for storage and how they acquire attitudinal and psychomotor skills (Ausubel, 1974; Craik & Lockhart, 1972; Duffy & Cunningham, 1996; Keller, 1983; Meyer, 1998; Sternberg, 1998). The pedagogical module should incorporate ideas from these learning theories to develop strategies to promote students’ learning. For example, according to constructivism, knowledge is constructed through experience and application. The pedagogical module must include strategies to allow learners to apply what they learn so that they can acquire personal meaning and, therefore, learn from the interaction.

How the Pedagogical Module and Students Interact

The pedagogical module performs the same functions as a tutor in a one-to-one tutoring situation. It interactswith the student in selecting problems to be solved, monitoring and analyzing the student’s performance, providing assistance upon request, and selecting remedial materials. Teaching methods are determined on the basis of diagnostic information obtained in the student modeling process (Tennyson & Park, 1987). Given expertise in the subject domain and a model of the student’s present level of expertise, the pedagogical module selects an efficient path through its knowledge representation to generate expert behavior by the user. Initial teaching strategies, based on a prototype or on a student’s previous performance, are modified as the student model evolves. The pedagogical strategies could be presenting increasingly complex concepts or problems, simulating phenomena, conducting Socratic tutoring with correction of student’s misconceptions, and modeling of expert problem solving via coaching.

Intervention in the Pedagogical Module

The fundamental issues for a tutor or coach are whether to intervene in the information flow during the learning process, what to discuss, which presentation strategies to use, and how much to say. In a learning-by-doing situation, intervention takes the form of an interruption by the coach when a systematic pattern of error has been spotted. In tutoring, intervention involves judging when to shift between imparting new information and debugging the student’s current conceptions (Dede, 1986). In tutorial guidance, it is sometimes more effective to let the student search for a certain time before any interruption (Wenger, 1987; Woolf, 1987). However, if a student is left all of the time with no intervention, he or she may get lost or stuck. One major instructional factor identified by intelligent tutoring systems researchers is the effect of discontinuous information flow on students’ interest. Too frequent interventions may destroy students’ initiative and decrease motivation to learn. Important criteria in choosing when to interrupt include relevancy of the material and the memory capacity of the student. The information provided by the intervention should be directed at a particular weakness, useful in the immediate situation, and demonstrably superior to the student’s misconception (Brown, Burton, & de Kleer, 1982). The question is how to do this. We need to examine how tutors in a one-to-one situation function, and duplicate this expertise in an intelligent tutoring system.

When an interruption is indicated in the system, the choice of what to emphasize becomes the next issue. Sometimes, choice of content involves more global issues than responding to a particular pattern of student error. For example, at frequent intervals, a review of what has been learned so far is a useful tutoring technique. If a student is struggling to master a learning-by-doing experience, reducing the overall level of difficulty by simplifying the task can enhance motivation, diagnosis, and remediation (Goldstein, 1982). Selecting which strategy to use in transferring knowledge from intelligent tutoring systems to student is another issue. Knowledge of the user’s learning style from the student model is an important criteria for selection, as is the choice between descriptive (textual) and depictive (graphic) representation (Dede, 1986). Higher-order cognitive skills (such as the ability to visualize what a program is doing) can be developed using instructional strategies with visuals (e.g., using a simulated, simplified machine) to illustrate a step-by-step programming procedure (du Boulay, O’Shea, & Monk, 1981).

The Socratictutor has been a powerful instructional model in intelligent tutoring systems. The mixed initiative—either the user or device may begin an interaction—combined with the Socratic model gives an opportunity to refine the student model. The goals of the Socratic tutor are to follow a domain- dependent script of knowledge, using production rules to identify student misconceptions and to offer counterexamples that build a new understanding (Dede, 1986) Another issue in the explanatory paradigm is in indicating student’s error, student’s self-image, and at the same time, motivation must not be lowered. The instructional response must be at an appropriate level of detail, not too specific to be boring and not too general that the student cannot use the information (Dede, 1986).

Extracting Pedagogical Knowledge

In a pedagogical module, one has to determine on what basis a tutor makes a decision in a one-to-one situation. What kind of knowledge does one use in the tutoring process? Two kinds of knowledge may be elicited from experts: (a) process knowledge, which is defined as the strategies and procedures used in problem solving; and (b) content knowledge, which represent the actual facts and rules used by experts in solving problems. The two kinds of knowledge cannot be divided into mutually exclusive classes (Garg-Janardan & Salvendy, 1987). Process knowledge includes the methods by which a subset of the content knowledge is accessed, combined, and used to solve problems. Combinations of content and process knowledge and heuristics may be used successfully so often that they become automated and stored as chunks by students. As soon as the student recognizes a pattern in a given problem, the associated chunk is executed. These chunks form as an individual’s experience and expertise level increase and render the process-content distinction more ambiguous. In addition to the process and content knowledge, tutors have to use pedagogical knowledge. How does the tutor combine these knowledge types in the tutoring process? Can they be separated, or should they be integrated? These are important research questions. Once the appropriate pedagogical strategy is determined for a learner, this must be presented to the learner through the interface module, which is customized for individual learners based on their styles and preferences.



 < Day Day Up > 



Designing Distributed Environments with Intelligent Software Agents
Designing Distributed Learning Environments with Intelligent Software Agents
ISBN: 1591405009
EAN: 2147483647
Year: 2003
Pages: 121

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net