Knowledge Cognitive theories and models tend to implicitly assume that a cognitive architecture-a... more Knowledge Cognitive theories and models tend to implicitly assume that a cognitive architecture-augmented with learning mechanisms-plus a task-specific knowledge base are sufficient to explain human cognition. Consequently, recent research has emphasized analogy and case-based reasoning, expertise, and the so-called situatedness of cognitive processes. These approaches focus primarily on the concrete and particular and ignore the abstract categories, methods, principles, schemas, and regulatory ideas that philosophers from Plato to Immanual Kant and Bertrand Russell assumed constitute the basic machinery of the human mind. Not everyone has jumped onto the particularity bandwagon. In Outsmarting IQ and elsewhere (Perkins & Salomon, 1989), David Perkins has championed the importance of what he calls reflective intelligence. To see how reflective intelligence can be both a source of generality and a source of individual differences, you need to consider some specific proposals as to the nature of the abstract knowledge structures that may underpin this type of intelligence. In Outsmarting IQ Perkins suggested that abstract knowledge can be conceptualized as a repertoire of dispositions (pp. 275-277). He proposed a list of seven thinking dispositions that characterize good thinking. These include the disposition to be adventurous in your thinking (e. g., to search far and wide for ideas and to be willing to consider seemingly weird solutions) and the disposition to be intellectually careful (e.g., to have a tendency to double check everything twice, to avoid fallacies, to ask for evidence, to probe for weaknesses, etc.). The seven thinking dispositions are abstract in character and have the potential to be useful in almost any domain or task. A different possibility is that reflective intelligence operates with a repertoire of abstract schemas (Ohlsson, 1993). Abstract schemas have many of the same properties as domain-specific schemas (internal structure, open slots), but the relations that connect the slots are so general as to be applicable across domains, and the slots have no constraints on their fillers. An abstract schema is pure structure, as it were. This content downloaded from 157.55.39.237 on Thu, 07 Jul 2016 04:53:57 UTC All use subject to http://about.jstor.org/terms
Differences in working memory capacity (WMC) relate to performance on a variety of problem solvin... more Differences in working memory capacity (WMC) relate to performance on a variety of problem solving tasks. High WMC is beneficial for solving analytical problems, but can hinder performance on insight problems (DeCaro & Beilock, 2010). One suggested reason for WMC-related differences in problem solving performance is differences in strategy selection, in which high WMC individuals tend toward complex algorithmic strategies (Engle, 2002). High WMC might increase the likelihood of nonoptimal performance on Luchins’ (1942) water jar task because high WMC solvers tend toward longer solutions, not noticing when shorter solutions become available. We present empirical data showing this effect, and a computational model that replicates the findings by choosing among problem solving strategies with different WM demands. The high WMC model used a memoryintensive strategy, which led to long solutions when shorter ones were available. The low WMC model was unable to use that strategy, and switc...
Based on our empirical studies of effective human tutoring, we developed an Intelligent Tutoring ... more Based on our empirical studies of effective human tutoring, we developed an Intelligent Tutoring System, iList, that helps students learn linked lists, a challenging topic in Computer Science education. The iList system can provide several forms of feedback to students. Feedback is automatically generated thanks to a Procedural Knowledge Model extracted from the history of interaction of students with the system. This model allows iList to provide effective reactive and proactive procedural feedback while a student is solving a problem. We tested five different versions of iList, differing in the level of feedback they can provide, in multiple classrooms, with a total of more than 200 students. The evaluation study showed that iList is effective in helping students learn; students liked working with the system; and the feedback generated by the most sophisticated versions of the system is helpful in keeping students on the right path.
Researchers argue that dissatisfaction with a misconception is a prerequisite for adopting an alt... more Researchers argue that dissatisfaction with a misconception is a prerequisite for adopting an alternative conception and that having clear feedback aids learning. The present study investigated the importance of ambiguity (having response options that support both the misconception and target learning category), falsification, and category induction opportunities when overriding a prior conception in favor of a new conception. The results suggest that ambiguity and direct falsification opportunities may aid in learning more than having both direct falsification and induction opportunities, which may be better than ambiguity and providing induction opportunities without direct falsification. Ambiguity may improve learning when coupled with falsification opportunities. Implications are discussed.
Traditional knowledge representations were developed to encode complete explicit and executable p... more Traditional knowledge representations were developed to encode complete explicit and executable programs, a goal that makes them less than ideal for representing the incomplete and partial knowledge of a student. In this paper, we discuss state constraints, a type of knowledge unit originally invented to explain how people can detect and correct their own errors. Constraint-based student modeling has been implemented in several intelligent tutoring systems (ITS) so far, and the empirical data verifies that students learn while interacting with these systems. Furthermore, learning curves are smooth when plotted in terms of individual constraints, supporting the psychological appropriateness of the representation. We discuss the differences between constraints and other representational formats, the advantages of constraint-based models and the types of domains in which they are likely to be useful.
Knowledge Cognitive theories and models tend to implicitly assume that a cognitive architecture-a... more Knowledge Cognitive theories and models tend to implicitly assume that a cognitive architecture-augmented with learning mechanisms-plus a task-specific knowledge base are sufficient to explain human cognition. Consequently, recent research has emphasized analogy and case-based reasoning, expertise, and the so-called situatedness of cognitive processes. These approaches focus primarily on the concrete and particular and ignore the abstract categories, methods, principles, schemas, and regulatory ideas that philosophers from Plato to Immanual Kant and Bertrand Russell assumed constitute the basic machinery of the human mind. Not everyone has jumped onto the particularity bandwagon. In Outsmarting IQ and elsewhere (Perkins & Salomon, 1989), David Perkins has championed the importance of what he calls reflective intelligence. To see how reflective intelligence can be both a source of generality and a source of individual differences, you need to consider some specific proposals as to the nature of the abstract knowledge structures that may underpin this type of intelligence. In Outsmarting IQ Perkins suggested that abstract knowledge can be conceptualized as a repertoire of dispositions (pp. 275-277). He proposed a list of seven thinking dispositions that characterize good thinking. These include the disposition to be adventurous in your thinking (e. g., to search far and wide for ideas and to be willing to consider seemingly weird solutions) and the disposition to be intellectually careful (e.g., to have a tendency to double check everything twice, to avoid fallacies, to ask for evidence, to probe for weaknesses, etc.). The seven thinking dispositions are abstract in character and have the potential to be useful in almost any domain or task. A different possibility is that reflective intelligence operates with a repertoire of abstract schemas (Ohlsson, 1993). Abstract schemas have many of the same properties as domain-specific schemas (internal structure, open slots), but the relations that connect the slots are so general as to be applicable across domains, and the slots have no constraints on their fillers. An abstract schema is pure structure, as it were. This content downloaded from 157.55.39.237 on Thu, 07 Jul 2016 04:53:57 UTC All use subject to http://about.jstor.org/terms
Differences in working memory capacity (WMC) relate to performance on a variety of problem solvin... more Differences in working memory capacity (WMC) relate to performance on a variety of problem solving tasks. High WMC is beneficial for solving analytical problems, but can hinder performance on insight problems (DeCaro & Beilock, 2010). One suggested reason for WMC-related differences in problem solving performance is differences in strategy selection, in which high WMC individuals tend toward complex algorithmic strategies (Engle, 2002). High WMC might increase the likelihood of nonoptimal performance on Luchins’ (1942) water jar task because high WMC solvers tend toward longer solutions, not noticing when shorter solutions become available. We present empirical data showing this effect, and a computational model that replicates the findings by choosing among problem solving strategies with different WM demands. The high WMC model used a memoryintensive strategy, which led to long solutions when shorter ones were available. The low WMC model was unable to use that strategy, and switc...
Based on our empirical studies of effective human tutoring, we developed an Intelligent Tutoring ... more Based on our empirical studies of effective human tutoring, we developed an Intelligent Tutoring System, iList, that helps students learn linked lists, a challenging topic in Computer Science education. The iList system can provide several forms of feedback to students. Feedback is automatically generated thanks to a Procedural Knowledge Model extracted from the history of interaction of students with the system. This model allows iList to provide effective reactive and proactive procedural feedback while a student is solving a problem. We tested five different versions of iList, differing in the level of feedback they can provide, in multiple classrooms, with a total of more than 200 students. The evaluation study showed that iList is effective in helping students learn; students liked working with the system; and the feedback generated by the most sophisticated versions of the system is helpful in keeping students on the right path.
Researchers argue that dissatisfaction with a misconception is a prerequisite for adopting an alt... more Researchers argue that dissatisfaction with a misconception is a prerequisite for adopting an alternative conception and that having clear feedback aids learning. The present study investigated the importance of ambiguity (having response options that support both the misconception and target learning category), falsification, and category induction opportunities when overriding a prior conception in favor of a new conception. The results suggest that ambiguity and direct falsification opportunities may aid in learning more than having both direct falsification and induction opportunities, which may be better than ambiguity and providing induction opportunities without direct falsification. Ambiguity may improve learning when coupled with falsification opportunities. Implications are discussed.
Traditional knowledge representations were developed to encode complete explicit and executable p... more Traditional knowledge representations were developed to encode complete explicit and executable programs, a goal that makes them less than ideal for representing the incomplete and partial knowledge of a student. In this paper, we discuss state constraints, a type of knowledge unit originally invented to explain how people can detect and correct their own errors. Constraint-based student modeling has been implemented in several intelligent tutoring systems (ITS) so far, and the empirical data verifies that students learn while interacting with these systems. Furthermore, learning curves are smooth when plotted in terms of individual constraints, supporting the psychological appropriateness of the representation. We discuss the differences between constraints and other representational formats, the advantages of constraint-based models and the types of domains in which they are likely to be useful.
Uploads
Papers by Stellan Ohlsson