A multidimensional scaling analysis of the development of animal names.

Investigated developmental changes in semantic structure using the INDSCAL multidimensional scaling procedure. Dissimilarity judgments of all possible parts of 10 animal names were obtained from 39 1st-, 3rd-, 6th-grade, and college students. The overall scaling solution revealed a semantic space consisting of the features of size, domesticity, and predativity. Further analyses revealed that this structure changes with development inasmuch as the perceptual feature of size becomes less salient and the more abstract features of predativity and domesticity becomes more salient. This finding is compared to those from earlier research in which other techniques were used, and it is concluded that multidimensional scaling is a useful tool for determining developmental changes in semantic structure. (16 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Assessing clinically relevant perceptual organization with multidimensional scaling techniques.

Multidimensional scaling (MDS) techniques provide a promising measurement strategy for characterizing individual differences in cognitive processing, which many clinical theories associate with the development, maintenance, and treatment of psychopathology. The authors describe the use of deterministic and probabilistic MDS techniques for investigating numerous aspects of perceptual organization, such as dimensional attention, perceptual correlation, within-attribute organization, and perceptual variability. Additionally, they discuss how formal quantitative models can be used, in conjunction with MDS-derived representations of individual differences in perceptual organization, to test theories about the role of cognitive processing in clinically relevant phenomena. They include applied examples from their work in the areas of eating disorders and sexual coercion. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A multidimensional scaling analysis of vowel discrimination in humans and monkeys

Multidimensional scaling (MDS) was used to compare perceptual maps for 10 synthetic English vowels in humans and Old World monkeys (Macaca fuscata and Cercopithecus albogularis). Subjects discriminated among the vowels using a repeating background procedure, and reaction times were submitted to an MDS analysis to derive measures of perceive similarity. The dimensions that emerged related to the frequencies of the first (F1), second (F2), and third (F3) formants. Human data indicated a good match to previous MDS studies using rating procedures or confusion matrices: The dominant dimension mapped onto vowel F2, the phonetically most important formant, and the second and third dimensions mapped onto F1 and F3, respectively. For monkeys, equal weightings occurred for F1 and F2, and F3 was not clearly represented. Monkey sensitivity to the formants appeared to relate to formant amplitudes. If monkeys are giving an accurate representation of the psychoacoustic relations among the formants, then our human results suggest that species-specific mechanisms, reflecting the salience of the phonetic feature of advancement, may contribute to vowel coding in humans.     

Relations between factor analysis and multidimensional scaling.

Describes the fundamental relations between multidimensional scaling and factor analysis. Metric and nonmetric versions of both models are described in terms of type of data analyzed, assumptions made, objectives, computational procedures, geometric representations of data and solutions, and psychological meaning of results. What is commonly taken to be a fundamental identity between the metric versions of the 2 models is shown to be merely the employment of the same theorems. The strongest relations between the techniques lie in the realm of individual differences models for multidimensional scaling. Several such models are presented and are shown to represent the application of the logic of factor analysis to the substance of multidimensional scaling. (17 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A multidimensional scaling analysis of schizophrenics' and normals' perceptions of verbal similarity.

Compared the judgments of similarity of words by 14 paranoid and 14 nonparanoid schizophrenics (mean ages, 29.43 and 30.64 yrs, respectively) with those by 14 normals (mean age, 32.79 yrs). The judgments were analyzed using an individual-differences multidimensional scaling procedure. A greater judgmental consistency was obtained among the normals than among the paranoid schizophrenics and larger differences among stimulus dimensions, in their contributions to predictable judgmental variance, were obtained for the normals as compared with the schizophrenics, especially the paranoid schizophrenics. Stimulus dimensions resembling the potency, activity, and evaluative dimensions of the semantic differential were less influential in the judgments of the schizophrenics, especially those classified as paranoid, than in the judgments of the normals. Results are discussed in terms of overinclusiveness among schizophrenics in multidimensional similarity judgments, as well as in terms of the potential importance to schizophrenics' deficiencies in semantic interpretations of the reduced influence of relevant dimensions of meaning. (37 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A comparison of linear and monotone multidimensional scaling models.

Multidimensional scaling solutions under the linear and monotone (metric and nonmetric) distance models were compared, in a simulation study, for several monotone and nonmonotone distortions. Data were generated from random configurations, over a wide range of conditions. Results indicate that, when its assumptions are met, the linear model performs best. When linearity assumptions are not met, the monotone model and the linear model applied to ranked data perform equally well. Recommendations based on these results are offered. (25 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Knowledge organization as a factor in career intervention outcome: A multidimensional scaling analysis.

Multidimensional scaling techniques (MDS) were used to examine the learning outcomes associated with J. L. Holland's Self-Directed Search (SDS; J. L. Holland, 1990). One hundred forty-six participants were randomly assigned to 3 conditions: the SDS, an attentional control, and a no treatment control. Based on the speculations of J. L. Holland, T. M. Magoon, and A. R. Spokane (1981), we hypothesized that participants who had completed the SDS would acquire Holland's complete RIASEC organizing scheme. No evidence of Holland's 2-dimensional RIASEC configuration was found for any of the groups. Examination of the 3-dimensional MDS solutions, however, suggested the presence of a prestige dimension and supported Holland's scheme as represented by D. J. Prediger's (1981) Things–People and Data–Ideas constructs. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Structure of causes for success and failure: A multidimensional scaling analysis of preference judgments.

Investigated whether B. Weiner's (1979) model of causal attributions applies to perceptions of the causes for success and failure. Instead of the usual similarity judgments, preference judgments were used to reveal the dimensions underlying these perceptions. Female subjects, randomly assigned to a success or failure condition, made preference judgments with regard to 12 causes for success or failure. Multidimensional scaling analysis uncovered internality, stability, and excusability dimensions, thus supporting Weiner's model at least partly. In addition, differences in the relative emphasis given to the dimensions were found between the success and failure conditions: Following success, the internality dimension was the most salient, whereas after failure the stability and excusability dimensions were the most important. Furthermore, after success subjects preferred more internal causes and more causes that seem inappropriate as an excuse. After failure, subjects chose more external and more excusable causes. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A multidimensional scaling analysis of supervisory dimensions and their perceived relevance across trainee experience levels.

The Level of Supervision Scale (LSS; R. D. Miars et al, 1983) was administered to 69 supervisors with experience supervising 1st-semester practicum and intern-level trainees. The structure of their responses to 28 LSS items assessing perceptions of supervision as they varied across these 2 trainee experience levels was examined using 3-way multidimensional scaling. A 4-dimensional solution was yielded, which accounted for 63% of the variance. The 4-dimensions are (1) enhancing dynamic understanding, (2) didactic instruction, (3) counseling vs supportive supervisory functions, and (4) authoritative vs collaborative style. The study demonstrated clear differences in the salience of the dimensions across supervisors' perceptions of the 2 trainee experience groups. Supervisors perceived the enhancing dynamic understanding dimension as most important in their supervision of interns, whereas didactic instruction was most salient in perceptions of supervising beginning trainees. No differences were found across experience level for the other 2 dimensions. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The analysis of job performance by multi-dimensional scaling techniques.

The application of multidimensional scaling methods to the analysis of job performance was explored. Men experienced in the job of Naval aviation electronics technician designated 18 tasks as constituting that job at the entry levels. Supervisory personnel then judged the similarity between all pairs of these tasks. The resulting scaled similarity estimates were analyzed by standard multidimensional scaling techniques. The work performed by aviation electronics technicians at the job entry levels was perceived by supervisors as involving 4 basic dimensions. It appears to be feasible and fruitful to apply multidimensional scaling techniques to the analysis of job performance. (22 ref.) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Application of multidimensional scaling to the analysis of sensory evaluations of stimuli with known attribute structures.

Examined sensory perceptions of dimensionally simple stimuli composed of 1 or 2 objectively measurable attributes of flavor (sweet, sour) and color (red) using multidimensional scaling analysis. In 3 experiments with 20–30 19–36 yr old Ss each, high correlations were found between the objectively measurable physical characteristics of the stimuli and the perceptual spaces obtained by multidimensional scaling analysis of expressed gustatory and visual perceptions. Correlations were highest for stimuli sets containing single attributes detected by either gustation or vision and decreased when more complex sensory attributes were presented. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Non-metric multidimensional scaling in the analysis of neuroanatomical connection data and the organization of the primate cortical visual system

Neuroanatomists have established that the various gross structures of the brain are divided into a large number of different processing regions and have catalogued a large number of connections between these regions. The connectional data derived from neuroanatomical studies are complex, and reliable conclusions about the organization of brain systems cannot be drawn from considering them without some supporting analysis. Recognition of this problem has recently led to the application of a variety of techniques to the analysis of connection data. One of the techniques that we previously employed, non-metric multidimensional scaling (NMDS), appears to have revealed important aspects of the organization of the central nervous system, such as the gross organization of the whole cortical network in two species. We present here a detailed treatment of methodological aspects of the application of NMDS to connection data. We first examine in detail the particular properties of neuroanatomical connection data. Second, we consider the details of NMDS and discuss the propriety of different possible NMDS approaches. Third, we present results of the analyses of connection data from the primate visual system, and discuss their interpretation. Fourth, we study independent analyses of the organization of the visual system, and examine the relation between the results of these analyses and those from NMDS. Fifth, we investigate quantitatively the performance of a number of data transformation and conditioning procedures, as well as tied and untied NMDS analysis of untransformed low-level data, to determine how well NMDS can recover known metric parameters from artificial data. We then re-analyse real connectivity data with the most successful methods at removing the effects of sparsity, to ensure that this aspect of data structure does not obscure others. Finally, we summarize the evidence on the connectional organization of the primate visual system, and discuss the reliability of NMDS analyses of neuroanatomical connection data.     

Using distance information in the design of large multidimensional scaling experiments.

Studied which distances are most important in determining the recovery performance of a nonmetric multidimensional scaling algorithm. Using Monte Carlo methods, it is shown that the large distances are critical to satisfactory performance, whereas the small and the medium distances play a much less crucial role. This finding has been reliably demonstrated across a variety of conditions, although only for a single combination of dimensionality and number of points. Parallels between this work and previous results obtained using cyclic and other incomplete designs are noted. On the basis of these results some recommendations to experimenters regarding data collection procedures are presented; these represent a simple alternative to the methods advocated by I. Spence and D. W. Domoney (see record 1975-10755-001). (8 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Quantifying categorical observations of social interactions through multidimensional scaling.

The use of correlational techniques for analyzing interpersonal interactions frequently requires the transformation of the observational data from a categorical to a quantitative form. The methods currently used for this transformation can drastically reduce the information of the data. The authors present a data-transformation method based on multidimensional scaling techniques that can both retain more of the complexity of the data and facilitate its examination. These procedures are applied to a family interaction coding system, and they identify the 2 dimensions of prosocial–deviance and high–low involvement. Results are applied to observations of triadic parent–child interactions. Behaviors of each interactant as directed toward each of the other interactants were continuously coded in 6-sec intervals for 40 min. Increased information about the behavior categories and the observed interactions is demonstrated. (28 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Empirical modeling of an alcohol expectancy memory network using multidimensional scaling.

Risk-related antecedent variables can be linked to later alcohol consumption by memory processes, and alcohol expectancies may be one relevant memory content. To advance research in this area, it would be useful to apply current memory models such as semantic network theory to explain drinking decision processes. Multidimensional scaling (MDS) was used to empirically model a preliminary alcohol expectancy semantic network, from which a theoretical account of drinking decision making was generated. Subanalyses showed how individuals with differing alcohol consumption histories may have had different association pathways within the expectancy network. These pathways may have in turn influenced future drinking levels and behaviors while the person was under the influence of alcohol. All individuals associated positive/prosocial effects with drinking, but heavier drinkers indicated arousing effects as their highest probability associates, whereas light drinkers expected sedation. An important early step in this MDS modeling process is the determination of iso-meaning expectancy adjective groups, which correspond to theoretical network nodes. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Application of scaling techniques to the evaluation of counseling outcomes.

A reasonably comprehensive review of the literature has revealed largely inadequate attempts to evaluate counseling, due chiefly to improper criterion development. A consolidation of various criterion issues reveals that multiple criteria and some combination of the other types of criteria discussed are most promising for application in research. A consolidation of scaling issues emphasizes multiple ratings, the closing gap between physical and psychological continua, and the improper use of theoretical definitions for criterion development. A review of rating scales covering the last 20 yr. reveals a general implementation of relevant criterion issues, but little attention to relevant scaling issues. It is concluded that each phase of rating scale development demands proper scaling techniques, i.e., factor analysis, and multidimensional scaling analysis. Other scaling methods are discussed, and item construction and analysis are developed briefly. (2 p. ref.) (PsycINFO Database Record (c) 2010 APA, all rights reserved)