Rachel Or-Bach

Educational benefits of metadata creation by students

Repositories of learning objects are developed and used by computer science educators especially for the teaching of programming, data types, and algorithms. Developing a learning object requires decisions about the granularity and the respective metadata for further retrieval. In this paper, we describe educational advantages of learning activities that deal with metadata creation by students. The learning objects relate to computer programming examples that the students use and reuse throughout a course.

Support options provided and required for modeling with DynaLearn--A case study

Science educators strongly advocate the importance of scientific modeling within science education. Although widely advocated for students, modeling is a complex task involving integration of topics, “languages” and abstraction levels. Thus support for the modeling task and for developing modeling skills is required. The goal of our study was to explore how novice modelers use several support options while performing modeling assignments with DynaLearn—an intelligent learning environment for qualitative modeling. The support options differ by the type of support, the presentation of the support, the relation to previous support, the adaptation to the learner and timing. Findings are expected to influence modifications and further development of support methods as well as providing guidelines for effective teaching using DynaLearn. Additional contributions of the study are insights on how novice modelers approach a modeling task, what type of support they are looking for, how they use each of the different support types, and what kind of instructional interventions might be required.

Pair Modeling with DynaLearn - Students' Attitudes and Actual Effects

With DynaLearn learners can construct scientific knowledge by manipulating icons and their inter-relationships, using a diagrammatic representation. The diagrams represent models that can be simulated, confronting learners with the logical consequences of the knowledge they expressed. Such modeling activities are highly advocated by science educators. Learning from the construction and debugging processes of modeling can be enhanced by collaboration. The modeling elements can serve as anchors for discussing, justifying, and explaining the model. Researchers have suggested various ways of supporting collaboration. In this study we employed Pair Modeling, which is an adaptation of the pair programming technique that is used for enhancing collaborative programming both in the industry and in academia. In this paper we present encouraging results for the use of this collaboration technique based on assignments’ scores, observations, and a questionnaire. Students’ attitudes were neutral on the average, but the average score of the group that employed Pair Modeling was significantly higher than the average score of the control group that employed unstructured pair collaboration. We discuss the implications of the obtained results and the limitations of the study.

Design considerations for supporting collaborative modeling

The design of learning environments that support collaborative learning involves several types of considerations. A basic consideration that affects most of the design decisions relates to the educational philosophy, which is strongly influenced by the learning theory the designer adopts. Most of the design decisions within a constructivist approach to learning entail various tradeoffs. We describe the considerations, tradeoffs and decisions involved in the design of support interventions for collaborative modeling. The description includes the interaction facilities and the design for the automatic generation of adaptive interventions.