Proceedings of the Annual Meeting of the Cognitive Science Society

Variation in second language acquisition is evident from earliest stages. This study examined effects of learning tasks (retrieval practice, comprehension, verbal repetition) on comprehension of Turkish as a new language. Undergraduates (N = 156) engaged with Turkish spoken dialogues in a computer-assisted language learning session via Zoom, with learning tasks manipulated between-subjects. Participants completed pre/posttests assessing comprehension of Turkish number and case marking, a vocabulary test, and open-response questions gauging explicit awareness. The retrieval-practice group showed highest performance overall, after controlling for significant effects of nonverbal ability and pretest. For comprehension of number/case marking, the comprehension group performed comparably to the retrieval-practice group. For vocabulary comprehension, the verbal-repetition group performed comparably to the retrieval-practice group. Differential performance associated with learning tasks indicates benefits of testing and production and aligns with transfer-appropriate processing. As predicted by the noticing hypothesis, explicit awareness of number and case marking correlated with comprehension accuracy.

The inference capabilities of humans suggest that they might be using algorithms with high degrees of parallelism. This paper develops a completely parallel connectionist inference mechanism. The mechanism handles obvious inferences, where each clause is only used once, but may be extendable to harder cases.

A model of computer-manual comprehension is proposed in which four processes operate simultaneously: task-mapping of the structure of regular procedures onto the structure of computer commands, constructing a mental model of the computer system, inducing the command language grammar, and learning the structure of computer procedures. Findings from a study of five novices' comprehension problems with UNIX documentation are analyzed in terms of these four processes. Two of the four processes -- task-mapping and procedure learning -- are described in this paper. The analysis focuses on the knowledge structures involved in comprehending a technical test.

The CEL model of learning and memory (Components of Episodic Learning) [Granger 1982,1983a,1983b] provides a process model of certain aspects of learning and memory in anumals and humans. The model consists of a set of asynchronous and semi-independent functional operators that collectively create and modify memory traces as a result of experience. The model conforms to relevant results in the learning literature of psychology and neruobiology. There are two goals to this work: one is to create a set of working learning systems that will improve their performance on the basis of experience, and the other is to compare these systems that will improve their performance on the basis of experience, and the other is to compare these systems' performance with that of living systems, as a step towards the eventual comparative characterization of different learning systems. Parts of the model have been implemented in the CEL-0 program which operates in a 'Maze-World' simulated maze environment. The program exhibits simple exploratory behavior that leads to the acquisition of predictive and discriminatory schemata. A number of interesting theoretical predictions have arisen in part from observation of the operation of the program, some of which are currently being tested in neurobiological experiments. In particular, some neurobiological evidence for the existence of multiple, seperable memory systems in humans and animals is interpreted in terms of the model, and some new experiments are suggested arising from the model's predictions

Very little of the knwoledge that an operator of a complex physical system brings to the job is purely quantitative in form. Virtually all of an operator's knowledge can be represented as qualitative relations or quasi-quantitative relations such as rough proportionalities. The realization that computer-based Instruction syste,s need to provide instructions and explanations In terms that students can use, that is, often in qualitative terms, has led to recent efforts in cognitive science and artificial intelligence to develop qualitative simulation models of complex dynamic systems. In this paper we discuss theoretical and pragmatic problems involved in using qualitative models to support automated explanation facilities.

Children (4 to 6 years old) were presented with problems requiring from 4 to 7 moves for solution. The problems were constructed such that the subgoal ordering was ambiguous. Children's performance was consistent with a generate and test strategy that had a 2-move lookahead for a goal state, a no-backup constraint, and some partial evaluation of progress toward the goal.

Most research into problem solving has considered each problem to be solved as a unique event. Our observations lead us to conclude that much of the problem solving perople do is based on previous experience. Analogy to previous experience contributes to the knowledge available for later problem solving. This paper presents a framework for those components of problem solving which rely on previous experience. The processes involved and the organization of experience which supports those processes are considered. Examples are drawn from two problem domains: diagnosis and treatment of modd disorders and plan selection for resolution of disputes.

This paper investigates two contexts of children's developing knowledge of the physical world: (1) the macro-context of different age cohorts (8-9 years versus 11-12 years); and (2) the micro-context of a one-hour experimental session. Twenty subjects were video-taped, constructing goal-states for a task inolving gears. Four distinct systemic approaches or problem spaces were identified: (a) Euclidean, (b) Kinematic, (c) Dynamic, and (d) Topological. The Arithmetic Modifier, effecting a numerical charaterization of a problem space, can operate on any of the four. Cross-age, there was the substantial overlap of inital problem space emplyed, and minimal overlap of a final problem space. This frequency of adaptive shift in problem space, strongly and positively correlated with age, suggests that, when confronted with an unfamiliar task domain, the capacity to recognize a problem space as inappropriate and to evoke another more adequate problem space appears to be a component of the answer to the classic question, "What develops?"

In this paper, we discuss the spectrum of on-line assistance ranging from passive, canned to active, user-customized. We discuss various of on-line assistance: interactive introductory tutorials, on-line help, and on-line manuals. We then describe two steps to make on-line assistance more intelligent: (1) inclusion and customization of examples in the information provided the user; and (2) integration of various aspects of on-line assistance like tutorials and help.

Counterplanning can be successfully used against most methods of resolving goal conflicts. However, if one's intentions are disguised by deception then an opposing actor will use incorrect counterplanning or possibly none at all. This paper describes two components in the creation of a deception, the deception type and the enablement type, the range of their possible values, and how the selction of each can be used to create different decepetions for the same situation.

These studies discuss two possible explanations for the selective effects observed in lexical ambiguity studies: one is selective inhibition and the other is attention. The two views make different predictions when a neutral target item is introduced between presentation of a homograph and a subsequent related target. The data show little signs of selective suppression, but they do suggest that attention may increase priming without producing selectivity.