Determining the Dimensionality of Multidimensional Scaling Representations for Cognitive Modeling

Multidimensional scaling models of stimulus domains are widely used as a representational basis for cognitive modeling. These representations associate stimuli with points in a coordinate space that has some predetermined number of dimensions. Although the choice of dimensionality can significantly influence cognitive modeling, it is often made on the basis of unsatisfactory heuristics. To address this problem, a Bayesian approach to dimensionality determination, based on the Bayesian Information Criterion (BIC), is developed using a probabilistic formulation of multidimensional scaling. The BIC approach formalizes the trade-off between data-fit and model complexity implicit in the problem of dimensionality determination and allows for the explicit introduction of information regarding data precision. Monte Carlo simulations are presented that indicate, by using this approach, the determined dimensionality is likely to be accurate if either a significant number of stimuli are considered or a reasonable estimate of precision is available. The approach is demonstrated using an established data set involving the judged pairwise similarities between a set of geometric stimuli. Copyright 2001 Academic Press.     

On the Complexity of Additive Clustering Models

Additive clustering provides a conceptually simple and potentially powerful approach to modeling the similarity relationships between stimuli. The ability of additive clustering models to accommodate similarity data, however, typically arises through the incorporation of large numbers of parameterized clusters. Accordingly, for the purposes of both model generation and model comparison, it is necessary to develop quantitative evaluative measures of additive clustering models that take into account both data-fit and complexity. Using a previously developed probabilistic formulation of additive clustering, the Bayesian Information Criterion is proposed for this role, and its application demonstrated. Limitations inherent in this approach, including the assumption that model complexity is equivalent to cluster cardinality, are discussed. These limitations are addressed by applying the Laplacian approximation of a marginal probability density, from which a measure of cluster structure complexity is derived. Using this measure, a preliminary investigation is made of the various properties of cluster structures that affect additive clustering model complexity. Among other things, these investigations show that, for a fixed number of clusters, a model with a strictly nested cluster structure is the least complicated, while a model with a partitioning cluster structure is the most complicated. Copyright 2001 Academic Press.     

The Processing of Coreference for Reduced Expressions in Discourse Integration

Three reading-time experiments in Chinese are reported that test contrasting views of how pronominal coreference is achieved. On the one hand, studies of reading time and eye tracking suggest that reduced expressions, such as the pronoun he, serve as critical links to integrate separate utterances into a coherent model of discourse. On the other hand, probe-word recognition studies indicate that full anaphoric expressions, such as a repeated name, are more readily interpreted than reduced expressions due to their rich lexical information, which provides effective cues to match the representation of the appropriate referent in memory. The results indicate that the ease of integrating the critical referent into a model of discourse is a function of the congruence of lexical, semantic, and discourse features conveyed by a syntactically prominent reduced expression within linguistic input. This pattern supports the view that a reduced expression is interpreted on-line and indeed plays a critical role in promoting discourse coherence by facilitating the semantic integration of separate utterances.     

The Korean diasporic experience: measuring ethnic identity in the United States and China

The measurement and development of ethnic identity of Koreans living in the United States and China were examined in the context of different migration experiences. Three hundred sixteen undergraduate college students were administered the Multigroup Ethnic Identity Measure (MEIM; J. S. Phinney, 1992). Separate principal-component analysis for each sample was performed to create cross-culturally equivalent Ethnic Identity and Other-Group subscales. Differential item function analysis revealed no cultural item bias among the retained MEIM items. Korean Americans had higher scores on the Ethnic Identity and Other-Group subscales and were more likely to be classified as bicultural than were Korean Chinese. Korean Chinese were more likely to be classified as assimilated, traditional, or marginalized in their cultural orientation.     

Do situational expectations produce situational inferenes? The role of future expectations in directing inferential goals

Little is known about the conditions that lead observers to adopt different inferential goals in the context of their everyday lives. Four studies examined whether future expectations created situational inferential goals. In 2 quasiexperimental studies, students made more situational inferences for targets in their expected future careers. In 2 experiments, situational expectations were manipulated, and participants made more situational inferences when they expected to be in the same situation as the target. This tendency was stronger when observers devoted minimal cognitive effort to their inferences, suggesting that when observers have situational expectations, making situational inferences is less effortful than making dispositional inferences.     

The isomorphic mapping hypothesis: evidence from Korean

This paper evaluates the relative merits of the trace deletion hypothesis, which attributes agrammatic comprehension difficulties to the loss of traces, and the isomorphic mapping hypothesis, which proposes that agrammatics have difficulty understanding sentences in which there is a nonisomorphic mapping between the syntactic representation and the corresponding event in the real world. (The two are isomorphic if the order of NPs reflects the place of entities in the event's 'action chain.' Since agents act on themes and transmit them to goals, the agent-theme-goal order is isomorphic with the corresponding event but the agent-goal-theme order is not.) The two hypotheses contrast in the predictions they make concerning goal-theme and theme-goal patterns in Korean: the TDH predicts degraded performance on the theme-goal structure (which is derived from the goal-theme structure); the IMH predicts difficulties on the goal theme structure, since it is nonisomorphic. Results of a comprehension study involving four Korean Broca's aphasics provide strong support for the IMH over the TDH.     

Suboptimality in human categorization and identification

Categorization and identification decision processes were examined and compared in 4 separate experiments. In all tasks, the critical stimulus component was a line that varied across trials in length and orientation, and the optimal decision rules were always complex piecewise quadratic functions. Evidence was found that identification is mediated by separate explicit and implicit systems. In addition, a common type of suboptimality was found in both categorization and identification. In particular, observers apparently approximated the piecewise quadratic functions of the optimal decision rules with simpler piecewise linear functions. A computational model, which was motivated by a recent neuropsychological theory of category learning, successfully accounted for this suboptimal performance in both categorization and identification. The model assigns a key role to the striatum and assumes the observed suboptimality was largely due to massive convergence of visual cortical cells onto single striatal units.     

Moore's Paradox And Self-Ascribed Belief

Moore's paradox arises from the logicaloddity of sentences of the form`P and I do not believe that P'or `P and I believe that not-P'. Thiskind of sentence is logically peculiarbecause it is absurd to assert it, although it isnot a logical contradiction. In this paperI offer a new proposal. I argue that Moore's paradox arises because there is a defaultprocedure for evaluating a self-ascribed belief sentence and one is presumptivelyjustified in believing that one believes a sentence when one sincerely assents to it.     

Transversal-Universals in Discourse Ethics: Towards a Reconcilable Ethics Between Universalism and Communitarianism

This paper discusses the possibility of an ethics of difference. It begins with an introduction to current poststructural and critical theories in order to show their significance for transcultural politics and ethics. Its theme is formulated in terms of the debate between the affirmation of ethical cognitivism cast in the form of universalism and the advocacy of moral skepticism in the mode of communitarianism. Distancing itself from the idea of universal morality, this paper attempts to respond to the challenge of both communitarians and deconstructionists in contemporary French poststructuralism. In the end, it argues for transversality in place of universality.     

The Factor Structure of the Beavers Interactional Scales Revisited

The present study addressed limitations of a recent confirmatory factor analytic study challenging the hypothesized factor structure of the Beavers Interactional Scales. That dataset (n = 97) was reanalyzed and two additional datasets (n = 95, n = 91) were explored. With all three samples, taken separately and together, the Beavers Systems Model, when operationalized in the BIS, was not supported. A body of evidence is accumulating which suggests that clinicians and researchers should be wary of the BIS.