Modelling argumentation and modelling with argumentation

This paper discusses the epistemological and methodological bases of a scientific theory of meaning and proposes a detailed version of a formal theory of argumentation based on Anscombre and Ducrot's conception. Argumentation is shown to be a concept which is not exclusively pragmatic, as it is usually believed, but has an important semantic body. The bridge between the semantic and pragmatic aspects of argumentation consists in a set of gradual inference rules, called topoi, on which the argumentative movement is based. The content of each topos is determined at the pragmatic level, while the constraints on the forms of the topoi attached to a sentence are determined at the semantic level. Applications and possible applications toartificial intelligence and to cognitive sciences are discussed. In particular, the gradual models used to account for argumentation are shown to be extremely promising for Knowledge management, a discipline which includes knowledge acquisition, knowledge representation, transmission of knowledge (communication, interfaces, etc.), knowledge production (decision help, reasoning, etc.). A first formal model is presented and discussed: it is shown in details how it accounts for most of the argumentative features of sentences containing but, little and a little, and how it can be extended to describe sentences containing other argumentative connectives. However, this model is shown to be too simple and to violate the compositionality principle, which is shown, in the first section, to bean important methodological principle for any scientific theory. After a detailed analysis of the possible reasons for this violation, an improved model is proposed and its adequacy is discussed.     

Coherent psychometric modelling with Bayesian nonparametrics

In this paper we argue that model selection, as commonly practised in psychometrics, violates certain principles of coherence. On the other hand, we show that Bayesian nonparametrics provides a coherent basis for model selection, through the use of a ‘nonparametric’ prior distribution that has a large support on the space of sampling distributions. We illustrate model selection under the Bayesian nonparametric approach, through the analysis of real questionnaire data. Also, we present ways to use the Bayesian nonparametric framework to define very flexible psychometric models, through the specification of a nonparametric prior distribution that supports all distribution functions for the inverse link, including the standard logistic distribution functions. The Bayesian nonparametric approach provides a coherent method for model selection that can be applied to any statistical model, including psychometric models. Moreover, under a ‘non‐informative’ choice of nonparametric prior, the Bayesian nonparametric approach is easy to apply, and selects the model that maximizes the log likelihood. Thus, under this choice of prior, the approach can be extended to non‐Bayesian settings where the parameters of the competing models are estimated by likelihood maximization, and it can be used with any psychometric software package that routinely reports the model log likelihood.     

Elastic solutions with arbitrary elastic inhomogeneity and anisotropy

An efficient numerical algorithm is proposed to accurately compute the elastic fields in two-dimensional (2D) or three-dimensional (3D) microstructures with arbitrary elastic inhomogeneity and anisotropy. It combines the equivalent inclusion method of Eshelby, the microelasticity theory of Khachaturyan, and the spectral iterative perturbation method of Hu and Chen. Its efficiency is compared with those of existing approaches in the literature. The method can be conveniently implemented in phase-field modeling of stress-dependent microstructure evolution and/or of mass/electrical transport.     

Caregiver interactions with autistic children

Caregiver interactions with young autistic children were contrasted with those involving caregivers and developmentally matched mentally retarded and normal infants. Caregivers of autistic children were similar to other caregivers in their responsiveness to child nonverbal communication bids and in their engagement in mutually sustained play. Caregivers of autistic children were similar to caregivers of mentally retarded children in their greater use of control strategies. However, these two groups of caregivers differed in the particular strategies they used to shape their children's behavior. Caregivers of mentally retarded children pointed to objects while caregivers of autistic children spent more time physically holding their children on task. Individual differences within the autistic sample indicated that caregivers regulated their children's behavior less and showed more mutual play and positive feedback to more communicatively able autistic children. These findings suggest that caregivers respond differentially to the specific deficiencies shown by their children.     

Cognitive representations of functional interactions with objects

Two studies addressed people’s knowledge about the movements underlying functional interactions with objects, when the interactions were described by simple verbal labels expressing environmental goals. In Experiment 1, subjects rated each action with respect to six dimensions: which portion of the limb moved, distance moved, forcefulness, effectors involved, size of the contact surface, and resemblance to grasp. Ratings were systematic and fell on two distinct underlying factors related to limb movement and effector (usually the hand) configuration. In Experiment 2, subjects sorted a subset of the actions by similarity of movement. Clustering and multidimensional scaling solutions indicated that the six initial dimensions contributed to similarity judgments, along with additional parameters. The results support the existence of cognitively accessible, but still relatively specific, representations of functional actions, with potential implications for motor and memory performance.     

Best linear composites with a specified structure

Least squares linear composites of predictors for estimating several criteria are derived, satisfying the restriction that the composites have an arbitrary specified intercorrelation matrix. These composites are compared with the usual unrestricted regression composites. An illustrative example is provided. The derivation depends on a general result, given in an appendix, about best-fitting orthonormal transformations.This research was done while the author was a Fellow at The Center for Advanced Study in the Behavioral Sciences. The investigation was supported by a Public Health Service fellowship, 1 F3 MH-28, 495-01 (PS), from the National Institute of Mental Health. Preparation of the paper was supported in part by Grant MH 07722 from the National Institute of Mental Health, Public Health Service, for work done at Carnegie-Mellon University.     

Modeling pedagogical interactions with machine learning

Zusammenfassung. Dieser Artikel beschreibt einen Ansatz zur Codierung und Analyse pädagogischer Interaktionen zwischen Lernenden und Experten im Bereich der Luftfahrt-Technik im Rahmen einer intelligenten tutoriellen Lernumgebung zur Fehlererkennung und -behebung (Sherlock 2). Es werden zwei prototypische maschinelle Lernverfahren dargestellt, welche bei der Analyse eines Corpus von Dialogprotokollen eingesetzt werden. Die Verfahren generieren der Diskursstruktur möglichst gut entsprechende Grammatiken. Die gefundenen Grammatiken führten zu neuen Einsichten hinsichtlich der Funktion bestimmter Sprechakte in tutoriellen Dialogen sowie zu klaren und präzisen Modellen von Erklärungsdialogen bei Diagnoseproblemen. Summary. This paper describes our approach to coding pedagogical interactions that took place between avionics students and domain experts in an ITS for electronic troubleshooting called Sherlock 2. We also describe prototype machine learning systems that we developed to learn grammars of discourse structure. The grammars revealed instructional functions of particular speech acts that we had not been aware of, and provided concise models of explanations common in diagnostic tasks.     

Mothers' interactions with temperamentally frustrated infants

A sample of 162 six‐month‐old infants was selected from a larger sample of 346 infants on the basis of mothers' report of their infants' temperament and a laboratory assessment of temperament. Infants were classified as easily frustrated or less frustrated and observed in several types of interactions with their mothers in the laboratory. Mothers completed several measures that indicated their level of parenting stress, psychological functioning, and marital adjustment. Maternal behavior with infants was coded along the dimensions of sensitivity, intrusiveness, and physical stimulation. Results indicated that maternal intrusiveness was related to infant temperament and that maternal physical stimulation was predicted by an interaction of infant temperament and mothers' perceived parenting stress. Implications of these findings for mother–infant interaction and subsequent child adjustment are discussed. ©2004 Michigan Association for Infant Mental Health.     

Fathers' interactions with their high-risk infants

Fathers' face-to-face interactions with their 4-month-old high-risk infants were compared to mothers' interactions with the same infants and to those interactions of fathers and mothers of normal infants. The high-risk infants were less attentive and less affectively responsive than normal infants, and their mothers were more active. Fathers of both high-risk and normal infants engaged in more game playing and laughed more frequently than mothers during interactions. Although the behaviors of the normal and high-risk infants differed, as did the behaviors of mothers interacting with them, the fathers engaged in similar amounts of activity, smiling, laughing and playing games with normal and high-risk infants. Fathers may be less disturbed than mothers by their high-risk infants' lesser responsivity.     

Logistic response models with item interactions

Items that are clustered according to shared content may violate the principle of conditional independence commonly used in item response theory. This paper investigates the capabilities of a logistic item response model in relation to locally dependent item responses. The model includes main effect and interaction parameters that are computed as linear functions of the latent trait. The paper explains the interpretation of the parameters, the maximum likelihood estimation algorithm, the information matrix and some results concerning parameter identifiability. The problem of over-fitting the data is addressed in a simulation study, and two real data examples are described to illustrate the approach, one from the context of a sample survey and the other from ability testing using testlets.     

Statistical analysis for finger tapping with a periodic external stimulus

In this paper we outline an experiment in which seven subjects attempted to tap in synchrony with a pacing stimulus with the interstimulus interval that varied in a cycle (T,T+dT,T, T-dT) repeating every fourth interval, where T represents an unexpected step change of an interstimulus interval. Conditions included three levels of T= (400, 500, and 600 msec.) with four levels of dT= 1, 3, 5, and 7% of T). The statistical analysis indicated that small changes in T lead to rapid matching of the interresponse intervals to the new interstimulus intervals even when the variation in the input is not apparent to subjects.     

Brightness scaling of periodic tones

Brightness is an attribute often used by musicians when describing timbral characteristics. It is related to the spectral distribution of energy, as is sharpness, studied by Zwicker (Psychoacoustics: Facts and Models, 1990). In the current work, subjects adjusted the spectral slope and thus the spectral centroid (SC) of one of a pair of sounds to make it twice as bright as the other, so as to build a perceptual scale. The ratio of SC required to double brightness is a little less than 2 and decreases as the SC of the tones increases. For these tones, the ratio of brightness is statistically different from the ratio of sharpness calculated from published models.     

Mathematical modelling of nucleation and growth of crystals with buoyancy effects

A complete analytical solution of the integro-differential model describing the nucleation of crystals and their subsequent growth in a binary system with allowance for buoyancy forces is constructed. An exact analytical solution of the Fokker-Planck-type equation for the three-parameter density distribution function is found for arbitrary nucleation kinetics. Two important cases of the Weber–Volmer–Frenkel–Zel’dovich and Meirs kinetics are considered in some detail. It is shown that the solute concentration decreases and the distribution function increases with increasing the melt supercooling (with increasing the depth of a metastable system). It is demonstrated that the distribution function attains its minimum at a certain size of crystals owing to buoyancy forces.     

Compression of environmental representations following interactions with objects

Previous work reveals that interacting with all objects in an environment can compress spatial memory for the entire group of objects. To assess the scope and magnitude of this effect, we tested whether interacting with a subset of objects compresses spatial memory for all objects in an environment. Participants inspected objects in one or two unmarked regions of space, then recalled the distances between pairs of objects from memory. One group of participants picked up objects in both regions, a second group picked up objects in one region and passively viewed objects in the other region, and a third group passively viewed objects in both regions. When participants manually interacted with objects, they recalled shorter object-pair distances throughout the environment. The magnitude of this effect was the same, regardless of whether participants interacted with all objects in the environment or just a subset of them. Together, these findings suggest that interacting with objects can compress environmental representations in memory, even when observers interact with a relatively small subset of objects.     

Stages-of-learning analysis of developmental interactions in memory,with illustrations from developmental interactions in picture-word effects

Although Age × Treatment interactions have been widely viewed as key results in theories of memory development, most examples of such results (e.g., Age × Treatment interactions in the effects of item concreteness, item elaboration, and organizational context) have been identified in list-learning experiments where children of different ages receive a small, fixed number of study-test cycles on the target list. A stages-of-learning analysis reveals that such designs confound a treatment's potential interactions with age and its potential interactions with stage of learning. The analysis also reveals specific situations in which such designs will produce converging, diverging, and crossover Age × Treatment interactions even though the treatment does not interact with age in any way. A series of experiments is reported in which a mathematical model that incorporates stages-of-learning distinctions is applied to the well-known interaction between age and item concreteness (pictures versus words) in cued recall. Although a diverging Age × Treatment interaction has been observed in previous research, it was found that there are at least four different interactions between the picture-word manipulation and age; (a) a diverging interaction at storage that operates throughout elementary school; (b) a diverging interaction localized within retrieval learning and operating during the first half of the elementary school; (c) a converging interaction localized within retrieval learning and operating during the second half of elementary school; and (d) a diverging interaction localized within retrieval performance and operating throughout elementary school.     

Audio-visual interactions with intact clearly audible speech

The effects of viewing the face of the talker (visual speech) on the processing of clearly presented intact auditory stimuli were investigated using two measures likely to be sensitive to the articulatory motor actions produced in speaking. The aim of these experiments was to highlight the need for accounts of the effects of audio-visual (AV) speech that explicitly consider the properties of articulated action. The first experiment employed a syllable-monitoring task in which participants were required to monitor for target syllables within foreign carrier phrases. An AV effect was found in that seeing a talker's moving face (moving face condition) assisted in more accurate recognition (hits and correct rejections) of spoken syllables than of auditory-only still face (still face condition) presentations. The second experiment examined processing of spoken phrases by investigating whether an AV effect would be found for estimates of phrase duration. Two effects of seeing the moving face of the talker were found. First, the moving face condition had significantly longer duration estimates than the still face auditory-only condition. Second, estimates of auditory duration made in the moving face condition reliably correlated with the actual durations whereas those made in the still face auditory condition did not. The third experiment was carried out to determine whether the stronger correlation between estimated and actual duration in the moving face condition might have been due to generic properties of AV presentation. Experiment 3 employed the procedures of the second experiment but used stimuli that were not perceived as speech although they possessed the same timing cues as those of the speech stimuli of Experiment 2. It was found that simply presenting both auditory and visual timing information did not result in more reliable duration estimates. Further, when released from the speech context (used in Experiment 2), duration estimates for the auditory-only stimuli were significantly correlated with actual durations. In all, these results demonstrate that visual speech can assist in the analysis of clearly presented auditory stimuli in tasks concerned with information provided by viewing the production of an utterance. We suggest that these findings are consistent with there being a processing link between perception and action such that viewing a talker speaking will activate speech motor schemas in the perceiver.     

A one-step method for modelling longitudinal data with differential equations

Differential equation models are frequently used to describe non-linear trajectories of longitudinal data. This study proposes a new approach to estimate the parameters in differential equation models. Instead of estimating derivatives from the observed data first and then fitting a differential equation to the derivatives, our new approach directly fits the analytic solution of a differential equation to the observed data, and therefore simplifies the procedure and avoids bias from derivative estimations. A simulation study indicates that the analytic solutions of differential equations (ASDE) approach obtains unbiased estimates of parameters and their standard errors. Compared with other approaches that estimate derivatives first, ASDE has smaller standard error, larger statistical power and accurate Type I error. Although ASDE obtains biased estimation when the system has sudden phase change, the bias is not serious and a solution is also provided to solve the phase problem. The ASDE method is illustrated and applied to a two-week study on consumers’ shopping behaviour after a sale promotion, and to a set of public data tracking participants’ grammatical facial expression in sign language. R codes for ASDE, recommendations for sample size and starting values are provided. Limitations and several possible expansions of ASDE are also discussed.     

Manipulation of two-dimensionally periodic stimulus patterns

An increasing number of visual studies require precise knowledge of the two-dimensional Fourier components of periodic spatial patterns. For patterns that are more complex than a simple sine-wave grating, however, the frequencies, magnitudes, phases, and even the orientations of these components are not always obvious from inspection of the pattern. Although the fast Fourier transform can provide a finite, sampled approximation to any desired spectrum, blind use of a packaged computer program sacrifices the intuitive insight of the analytic transform, which is particularly important for conceiving new experiments. This paper draws attention to some analytic techniques involving arrays of Dirac delta functions, which the author has found helpful in designing and understanding both real andGedanken experiments in pattern perception.     

Assessment of parents' interactions with their developmentally delayed infants

An instrument is described that is used to assess interaction style in parent-infant dyads in which the infant is handicapped or developmentally delayed. The rating scales were designed to be easy to use by professionals and trained nonprofessionals, to be applicable to children who are functioning between birth and 18 months developmental age, and to involve a minimal time commitment so that they could be used in the context of early intervention programs. The rating scales cover intrusive, responsive, and general social behavior by the parent and social and avoidant behavior by the child. Successful implementation of these scales is described.