Aemilian Hron

Implicit and explicit dialogue structuring in virtual learning groups.

BACKGROUND The promotion of virtual learning groups by instructional means is gaining importance with the spread of telematic learning. In virtual learning groups, students discuss certain learning materials or cooperate in problem solving by means of computer-mediated communication. Due to the specific features of electronic communication, supportive means appear to be necessary for such learning situations. AIMS The investigation focuses on testing two different forms of dialogue structuring in virtual learning groups, in which students discuss a given subject matter together and clarify problems in understanding by means of synchronous computer-mediated communication. SAMPLE Forty-five students from various disciplines at the University of Tübingen, except students of physics, participated in the investigation. METHOD An experiment was conducted involving three-person groups in a telematic setting with two different kinds of dialogue structuring: implicit structuring induced group discussion on the subject matter by working on key questions in a preceding learning phase, whilst explicit structuring provided additional rules for discussion. These rules prompted group members to argue and aimed at directing them to equal participation. RESULTS Compared to a control group, both implicit and explicit structuring showed greater orientation on the subject matter and showed less off-task talk. Moreover, explicit structuring led to more coherence in subject matter discussion with regard to the completion of topics. However, post-test performance showed no significant difference in knowledge over that of the control group. CONCLUSION The results of the investigation show that dialogue structuring can be an adequate pedagogical approach for virtual learning groups.

Structural Inferences in Reading and Listening

Abstract Cognitive processing of instructional texts places special demands upon the recipient. The thematic macrostructure must be extracted through the formation of structural inferences. It is assumed that the formation of such structural inferences in listening to or reading an instructional text can be promoted differently due to the differences in the prerequisites for perception. An experimental study examined the effect of text accentuation measures upon the promotion of structural inferences. The effect of two variables was analysed: The original version of an instructional text from the Funkkolleg “The law” was compared against the same version in which the text - immanent macrostructure was highlighted with the aid of a signaling method. Both text versions were presented visually and orally. The promoting effect of the text accentuating measures was examined in terms of a presentation modus. Generally it was demonstrated that processing performance with written texts was significantly greater than with oral texts. The results specifically point out that signaling when used in texts for reading has a greater promoting effect than when used in texts for listening. The results suggest that the possibilities for the utilization of such structural aids as signaling with regard to listening are more limited or at least are different than on reading. The results make it obvious that when producing “written texts for reading” and “written texts for listening” one has to take into account the specific perception prerequisites for oral or visual presentation.

Manipulable graphics for computer-supported problem solving

The use of computer graphics for problem solving was investigated during use by individual learners and by distant pairs collaborating by interaction through the screen. In both investigations learners had to correct computer graphics representing population models, either by graphical manipulation or by written explanation. In both situations positive effects of manipulable graphics on problem solving performance were found. However, a detailed analysis of the interaction processes in the collaborative situation showed that, comparing both experimental conditions, static graphics led to a higher proportion of correct content-related statements and explanations of system relations. No differences were observed between both conditions on the participation structure within the learning dyads. The results suggest that for the effective use of graphics in computer-supported collaborative learning, structural support devices that encourage adequate processing and focus learners on the problem solving process seem to be necessary.

Metaphors as didactic means for multimedia learning environments

SUMMARY Increasingly, metaphors are becoming important design elements in multimedia learning environments. The didactic potential of metaphors when applied to multimedia learning systems lies in their ability to structure information according to theme and content. They can also facilitate navigation and arouse attention and interest in the subject matter. In order to foster the didactic use of metaphors, a model is developed for levels of metaphor mapping. This is oriented towards a cognitive engineering approach for describing aspects of user interfaces. With respect to this model, the didactic dimensions of width and depth of metaphors are differentiated. It is shown to what extent these dimensions should be taken into account in metaphor conceptions useful for learning in a multimedia environment, ie, structure metaphors and interaction metaphors, in order to enhance learning. Finally, the perspective of hyperworld metaphors is introduced as a useful metaphor conception for the future.

A Rule-Based Diagnosis System for Identifying Misconceptions in Qualitative Reasoning in the Physical Domain “Superposition of Motion”

An automatic diagnosis system for the assessment of conceptual knowledge with a focus on individual misconceptions was developed. The system is a central component of a computer-based learning environment on the physical domain superposition of motion. The diagnosis system is based on a production system approach and has been implemented as a classifier system. Conceptual knowledge manifesting itself in qualitative reasoning on physical problems concerning superposition of motion is represented by means of qualitative rules, the rules representing changes of physical parameters in qualitative terms. Processes of knowledge acquisition are modelled by means of strengthening and discrimination algorithms. The diagnosis sytem is suitable to generate automatically both the correct solution and incorrect solutions corresponding to the learner’s correct or incorrect qualitative reasoning. It compares these solutions with the learner’s actual solution and thus infers his or her respective conceptual knowledge. The system is able to discriminate between misconceptions resulting in identical problem solutions by means of conflict resolution procedures. To some extent it is also able to cope with ambiguous learner responses and noisy data attributable to unsystematic behavior. The diagnosis system has been implemented in LOOPS and is running on a Xerox 1108 workstation.

Cognitive Theories as a Basis for Student Modelling

Starting out from Anderson’s concept of procedural learning and White and Frederiksen’s approach of mental models, student modelling in the domain of intelligent tutoring systems is discussed with regard to cognitive approaches. Anderson’s approach provides a conceptual basis for analyzing knowledge acquisition processes with regard to cognitive skills. Within the framework of this approach Anderson models problem-solving behaviour and knowledge acquisition processes based on the description of an ideal student behaviour on the one hand (expert model), and incorrect behaviour on the other hand (bug library), applying production rules. White and Frederiksen focus primarily on conceptual knowledge and relate to the construct of mental models. Knowledge acquisition processes are modeled as a progression of mental models which the student passes through in the course of acquiring expertise. A more critical perspective reveals that Anderson lays emphasis on procedural learning, completely neglecting cognitive processes. The weak points of White and Frederiksen’s approach consist in the fact that knowledge acquisition processes are not described in psychological terms, but rather in physical terms. Although the above-mentioned ways of approaching the matter of knowledge acquisition differ widely, cognitive approaches allow a more differentiated analysis of the representation and acquisition of knowledge, thus providing an important basis for student modelling.

Computer-Based Learning Environment and Automatic Diagnosis System for Superposition of Motion

In the context of a computer-based learning environment on superposition of motion, an automatic diagnosis system on knowledge acquisition and on misconceptions was developed. The learning environment is designed according to the principle of inductive learning. It consists of a sequence of sixty similar tasks increasing in task demand and tutorial measures, involving simulations of the physical phenomenon. The diagnosis system is enabled to generate automatically both the correct solution and incorrect solutions which are based on the learner’s possible misconceptions. It compares these solutions with the learner’s actual solution and thus infers on his or her respective conceptual knowledge. The diagnosis system is rule-based and implemented as a classifier system. It uses both a strengthening algorithm and a discrimination algorithm. Learning environment and diagnosis system are implemented in LOOPS on a Xerox 1108 workstation.

Lernplattformen in der Schule

Lernplattformen sind webbasierte Softwaresysteme, die eine Reihe von Funktionalitaten zur Unterstutzung von Lehr-/Lernprozessen sowie entsprechende Lehrmaterialien (E-Learning Ressourcen) umfassen. Zugriff haben – ggf. mit differenzierten Rechten – sowohl Lehrende als auch Lernende oder sonstige in das Lerngeschehen involvierte Personen. Die Funktionalitaten sind je nach System in unterschiedlichem Ausmas und unterschiedlicher Differenziertheit gegeben. Als Kernfunktionalitaten anzusehen sind Benutzerverwaltung (z.B. Anmeldung mit Passwortzuteilung), Kursverwaltung (z.B. Einrichten von Kursen oder Gruppen, Zuordnung von Inhalten, Verwaltung von Gruppen-Daten), Rollen- und Rechtevergabe, Kommunikationsmethoden (z.B. Chat, Foren) und Werkzeuge fur das Lernen (z.B. Whiteboard, Notizbuch, Annotationen, Kalender) (vgl. Schulmeister 2003). Die den Lernenden zur Verfugung gestellten Lehrinhalte konnen entweder selbst erstellt (mit geeigneten Autorenwerkzeugen) oder hinzugekauft sein.

Simultaneous Processing of Different Problem Aspects in Expert Problem Solving

Expert problem solving in the physical domain is analyzed with respect to simultaneous processing of different problem aspects. On the basis of formal theories of commonsense knowledge in Artificial Intelligence a rational task analysis of a given physical problem leads to a conceptual model of problem solving. This model is a template for an empirical taks analysis by means of analyzing the verbal protocols of four experts. The study reveals that only in one of the experts, the verbal statements could be interpreted to a larger extent in the sense of the information processing searched for. It is assumed that simultaneous processing of different problem aspects leads to a higher redundancy in expert problem solving and thus brings about the chance for mutual control and testing of the single operations.

Ein Regel-Basiertes Diagnose-System zur Identifizierung von Fehlkonzepten

Fur ein computergestutztes Lehrprogramm zur physikalischen Bewegungsuberlagerung wurde ein Diagnosesystem entwickelt. Es ermoglicht die automatische Diagnose von korrektem und fehlerhaftem Wissen, das ein Lerner beim Losen physikalischer Probleme zeigt. Zu jeder Problemstellung generiert das System sowohl die richtige Losung, als auch falsche Losungen, die auf moglichen Fehlvorstellungen des Lerners uber den physikalischen Sachverhalt beruhen. Durch den Vergleich dieser Losungen mit dem Losungsergebnis des Lerners schliest das System auf das jeweilige Konzeptwissen des Lerners.