Jill H. Larkin

Expert and Novice Performance in Solving Physics Problems

Although a sizable body of knowledge is prerequisite to expert skill, that knowledge must be indexed by large numbers of patterns that, on recognition, guide the expert in a fraction of a second to relevant parts of the knowledge store. The knowledge forms complex schemata that can guide a problems interpretation and solution and that constitute a large part of what we call physical intuition.

Models of Competence in Solving Physics Problems

We describe a set of two computer-implemented models that solve physics problems in ways characteristic of more and less competent human solvers. The main features accounting for different competences are differences in strategy for selecting physics principles, and differences in the degree of automation in the process of applying a single principle. The models provide a good account of the order in which principles are applied by human solvers working problems in kinematics and dynamics. They also are sufficiently flexible to allow easy extension to several related domains of physics problems.

Teaching General Learning and Problem-Solving Skills.

This article describes the investigation and teaching of two general cognitive skills important in introductory physics. We first analyzed the various abilities needed for understanding a relation (definition or law) well enough to use it appropriately. Then we developed two different instructional methods for teaching students the general learning skill of gaining such an understanding of any new relation. We further taught students a simple strategy for problem solving. Our results indicate that students can indeed be taught such general cognitive skills and that they can transfer these skills to areas outside of physics.

Cognition of learning physics

Experienced physicists solve even simple textbook problems in ways that are very different form the solutions produced by beginning students. This paper describes these differences using a computer‐implemented model that ’’learns.’’ This research is set in the context of modern information‐processing psychology, and is related to other work relevant to how people know and learn in quantitative domains.

FERMI: A Flexible Expert Reasoner with Multi‐Domain Inferencing

Expert reasoning combines voluminous domain-specific knowledge with more general factual and strategic knowledge. Whereas expert system builders have recognized the need for specificity and problem-solving researchers the need for generality, few attempts have been made to develop expert reasoning engines combining different kinds of knowledge at different levels of generality. This paper reports on the FERMI project, a computer-implemented expert reasoner in the natural sciences that encodes factual and strategic knowledge in separate semantic hierarchies. The principled decomposition of knowledge according to type and level of specificity yields both power and cross-doman generality, as demonstrated in FERMIs ability to apply the same principles of invariance and decomposition to solve problems in fluid statics, DC-circuits, and centroid location. Hierarchical knowledge representation and problem-solving principles are discussed, and illustrative problem-solving traces are presented.

Novice Strategies for Processing Scientific Texts.

Three studies examined text features used by novice scientists to determine what is important in scientific texts. Expert and novice physicists selected the 10 most important sentences from physics texts, rated the importance of text sentences, and wrote summaries of physics texts. These data were analyzed to determine (1) the degree to which information type (i.e., definitions, facts, equations, and elaborations) was used as an indicator of importance, (2) the extent to which novices agreed with experts on what was important, and (3) how information type interacted with passage structure (i.e., level in passage hierarchy) in determining importance judgments. Both experts and novices considered definitions to be more important than facts but novices applied this rule more strongly. Both groups considered information high in the passage hierarchy to be more important than lower level information, but hierarchical level had a greater effect on the perceived importance of definitions than facts.

Novice rules for assessing importance in scientific texts

Abstract Previous research shows that content area novices judge certain categories of information (e.g., definitions, facts, equations) as more important than others. The current research tested the hypothesis that novice importance judgements are based on category membership, rather than content differences between categories. Subjects of varying expertise judged the importance of sentences in physics texts when they were presented in one of two forms: definitions or facts (Experiment 1), and equations or their verbal equivalents (Experiment 2). The two sentence versions were always identical in substantive content. Experts and naive subjects (subjects without physics training) judged these variants to be similar in importance. However, beginning physics students judged definition and equation versions as more important. Thus beginning-level students develop rules specifying what categories of information are important, so that sentence category is a salient text feature. Sentence category is irrelevant for experts, who judge importance according to content, and naive subjects, who have not formed expectations regarding the importance of information categories. These results suggest how a content schema might evolve in novice learners.

Attentional strategies for studying scientific texts

Content-area novices develop rules as to what types of information (e.g., definitions, facts, equations) are important in texts (Dee-Lucas & Larkin, 1988). The study reported here indicates that these rules influence text learning. Experts and novices read and recalled science texts. Reading times and recall for different types of content were compared for the two groups. Results indicate that novices’ importance rules function as part of novices’ control schema during reading, influencing their attentional processes and the resulting representation formed for the text. This was evident in qualitative differences between experts and novices in their attentional patterns and text recall. The study also found that the number of passage readings and the passage topic have a greater influence on the reading times of exports, but both groups adjust processing time according to the hierarchical level of the passage content. The findings are discussed in terms of their implications for novices learning from texts.

Organization and Comprehensibility in Scientific Proofs, or "Consider a Particle p....".

Scientific texts commonly present principles by first giving a proof and only afterward stating the principle-a «proof-first» organization. This specialized text structure differs from conventional structures in that it lacks thematic information to guide text processing. The current research examined the effects on comprehension of this proof-first organization. This was done by comparing the processing of proof-first texts to that of «principle-first» texts, in which the theme (i.e., the principle) is stated at the beginning

New tools for courseware production

PRODUCING SIGNIFICANT QUANTITIES of high-quality computer-based instructional materials is a major challenge. Using a computer as a medium for education is one of the most technically demanding man-machine interactions requiring not only calculational capabilities but also involving interactive graphics, response analysis, and complex sequencing. Moreover, pedagogical and esthetic aspects are much less well understood in this new medium than in older educational media such as books and lectures. We will attempt to provide a framework for understanding the organizational and technical issues associated with courseware production for university courses. We will focus on the production of university courses in which the computer plays a major but not the only role. We consider courses in which students typically spend several hours per week in computer-based activities but may also have lectures, discussions, textbooks, etc. Producing such substantial amounts of computer-based material is quite different from writing a few short programs to be used occasionally as adjuncts to a course, and careful attention must be paid to organizational and technical issues.First we will contrast the advantages and disadvantages of production by teams and by individuals. We favor making it technically and organizationally feasible for individual faculty members to write their own materials if they wish. If the technical tools are good enough to make this feasible, teams can also be more productive. The second part of this paper describes and contrasts existing tools including “formatters, ” non-programming authoring systems, standard programming languages, and special-purpose educational programming languages.