This paper presents a skill learning model CLARION. Different from existing models of mostly high-level skill learning that use a top-down approach (that is, turning declarative knowledge into procedural knowledge through practice), we adopt a bottom-up approach toward low-level skill learning, where procedural knowledge develops first and declarative knowledge develops later. Our model is formed by integrating connectionist, reinforcement, and symbolic learning methods to perform on-line reactive learning. It adopts a two-level dual-representation framework (Sun, 1995), with a combination of localist and distributed representation. We compare the model with human data in a minefield navigation task, demonstrating some match between the model and human data in several respects.

3 experiments were designed to demonstrate that classifying new letter strings as grammatical (i.e., conforming to a set of rules called a synthetic grammar) or ungrammatical may proceed from fragmentary conscious knowledge of the bigrams constituting the grammatical strings displayed in the study phase, rather than from an unconscious structured representation of the grammar, as Reber (1989) contended. In Experiment 1, grammaticality judgments of subjects initially studying grammatical letter strings did not differ from judgments by subjects learning from a list of the bigrams making up these strings. In Experiment 2, judgments about nongram-matical strings composed of valid bigrams placed in invalid locations were extremely poor, although better than chance. In Experiment 3 the explicit knowledge of bigrams as assessed by a recognition procedure appeared sufficient to account for observed performance on a standard test of grammaticality. A widely held model of cognition endows human subjects with the ability to implicitly abstract the regularities or high-level rules embodied in richly structured stimulus domains. Over the last 20 years, this general model has received strong

This paper is concerned with modeling the nature of expertise and its role in society in relation to research on expert systems and enterprise models. It argues for the adoption of a collective stance in which the human species is viewed as a single organism recursively partitioned in space and time into sub-organisms that are similar to the whole. These parts include societies, organizations, groups, individuals, roles, and neurological functions. Notions of expertise arise because the organism adapts as a whole through adaptation of its interacting parts. The phenomena of expertise correspond to those leading to distribution of tasks and functional differentiation of the parts. The mechanism is one of positive feedback from parts of the organism allocating resources for action to other parts on the basis of those latter parts past performance of similar activities. Distribution and differentiation follow if performance is rewarded, and low performers of tasks, being excluded by the f...

The relationship between implicit memory and implicit learning is explored. Dienes and Fahey (1995) showed that learning to control a dynamic system was mediated by a look-up table consisting of previously successful responses to specific situations. The experiment reported in this paper showed that facilitated performance on old situations was independent of the subjects' ability to recognize those situations as old, suggesting that memory was implicit. Further analyses of the Dienes and Fahey data replicated this independence of control performance on recognition. However, unlike the implicit memory revealed on fragment completion tasks, successful performance on the dynamic control tasks was remarkably resilient to modality shifts. The results are discussed in terms of models of implicit learning and the nature of implicit memory.

This paper describes an approach in the area of complex problem solving based on linear structural equations, called DYNAMIS. It can be seen as an attempt to establish an experimentally and systematically oriented line of research on complex problem solving which should overcome some of the problems of early research.

In developing autonomous agents, one usually emphasizes only (situated) procedural knowledge, ignoring more explicit declarative knowledge. On the other hand, in developing symbolic reasoning models, one usually emphasizes only declarative knowledge, ignoring procedural knowledge. In contrast, we have developed a learning model Clarion, which is a hybrid connectionist model consisting of both localist and distributed representations, based on the two-level approach proposed in Sun (1995). Clarion learns and utilizes both procedural and declarative knowledge, tapping into the synergy of the two types of processes, and enables an agent to learn in situated contexts and generalize resulting knowledge to different scenarios. It unifies connectionist, reinforcement, and symbolic learning in a synergistic way, to perform on-line, bottom-up learning. This summary paper presents one version of the architecture and some results of the experiments. Key Words: hybrid models, sequential decision ...

With the advent of computers in the experimental labs, dynamic systems have become a new tool for research on problem solving and decision making. A short review of this research is given and the main features of these systems (connectivity and dynamics) are illustrated. To allow systematic approaches to the influential variables in this area, two formal frameworks (linear structural equations and finite state automata) are presented. Besides the formal background, the article sets out how the task demands of system identification and system control can be realised in these environments, and how psychometrically acceptable dependent variables can be derived. The use of computer-simulated scenarios in problem-solving research has become increasingly popular during the last 25 years (for a representative collection of papers see, e.g., the two editions from Sternberg & Frensch, 1991, and Frensch & Funke, 1995). This new approach to problem solving seems attractive for several reasons. In contrast to static problems, computer-simulated scenarios provide a unique opportunity to study human problem-solving and decision-making behaviour when the task environment and subjects ’ actions change concurrently. Subjects can manipulate a specific scenario via a number of input variables (typically ranging from 2 to 20, and in some exceptional instances even up to 2000), and they observe the system’s state changes in a number of output variables. In exploring and/or controlling a system, subjects have to continuously acquire and use knowledge about the internal structure of the system.

The present study discusses findings that replicate and extend the original work of Burns and Vollmeyer (2002), which showed that performance in problem solving tasks was more accurate when people were engaged in a non-specific goal than in a specific goal. The main innovation here was to examine the goal specificity effect under both observation-based and conventional action-based learning conditions. The findings show that goal specificity affects the accuracy of problem solving in the same way, both when the learning stage of the task is observationbased and when it is action-based. Additionally, the findings show that, when instructions do not promote goal specificity, observation-based problem solving is as effective as action-based problem solving. 2 Keywords: Problem solving; Skill acquisition and learning; Observation vs. intervention

Using the Pathfinder Scaling Algorithm (Pathfinder) (Interlink, 1994), the researchers conducted a cognitive task analysis expertise in Human Performance Technology (HPT). The study investigated: 1) the extent to which Pathfinder-derived coherence scores were associated with other measures of HPT expertise; 2) how HPT experts organize their knowledge of the discipline; and 3) how experts organize their HPT knowledge differently than novices. Findings include: 1) a significant correlation between coherence and the number of HPT-related books participants had written; 2) a Pathfinder-derived concept map of HPT; and 3) expected novice/expert differences in Pathfinder similarity and relatedness scores. 2

The role of action has been strongly emphasized, not only in cognitive research on learning and problem solving, but also in education and instructional psychology. The Constructivism tradition has long asserted that action plays a crucial role for learners in constructing their own knowledge. In an educational context, active engagement entails students examining their own ideas, considering alternative explanations for newly taught concepts, and evaluating competing perspectives. Some theorists (e.g., Anzai & Simon, 1979) propose that these processes are found when learning is by doing. However, a constructivist perspective implies that instructional formats enable self-monitoring (e.g., Covington, 2000; Pintrich & De Groot, 1990), which includes reflective activities such as describing, explaining, and evaluative thinking (e.g., Covington, 2000; Zimmerman, 1990), which are not exclusive to action. The present article discusses findings that concern two related and thus far, unexplored two questions: How affective is observation-based learning in a complex skill learning task that usually requires processes that involve active engagement with it? How does monitoring affect the transfer of problem solving ability in complex skill learning task? The first aim of the article is to introduce ways of using common educational tools like the self-observation technique, which involves re-exposing individuals to their own self-generated behaviors, in novel ways that can provides insight into how people use self-regulatory mechanisms like monitoring on internally represented behaviors. The second aim is provide support for the view that in the absence of active learning, learning indirectly (i.e. Observation-based learning) is a practical and in some cases necessary method of knowledge and skill acquisition, and does not in turn lead to decrements in acquired knowledge and skill. Finally, the article presents the argument that the degree of self-monitoring that takes place may be a mediating factor in preserving the view that action has a special status in knowledge acquisition.