UNIDIMENSIONAL SIMILARITY REVISITED

It is argued that the use of direct scaling methods to derive the perceptual magnitudes in the similarity function is inadvisable on two grounds. (1) Attention may shift from one way of rating to another and thereby affect the perceptual magnitude so that what is rated may depend on the method of rating. (2) Direct methods of scaling seldom give consistent data and present knowledge of the reason for this is limited. The preferred procedure then is to study solely the structure of the similarity matrices themselves. When this is done on four sets of published data it is found that the similarity mechanism of Eisler and Ekman gives a close fit.     

Asymmetries of comparison

Tversky’s (1977) seminal work on directionality in judgments of similarity demonstrated that people may not judge the similarity of A to B to be equal to the similarity of B to A. In a series of studies, we investigated comparison asymmetries. In Experiment 1, our aim was to extend Tversky’s findings to a heterogeneous stimulus set, but no reliable asymmetries were observed. Experiment 2 employed a variety of comparison judgments, and, although some of these measures showed asymmetries, we still failed to observe asymmetries in rated similarity. A final attempt to obtain asymmetries used direction as a within-subjects factor, and for the first time, rating asymmetries were observed. Our data reinforce the idea of comparison asymmetries but suggest that similarity rating asymmetries are only observed under quite circumscribed conditions.     

Salience of similarity awareness and attraction: A comparison of balance vs reinforcement predictions

Reinforcement and balance theory are both viable explanatory contenders for attraction processes. Differential predictions were derived and tested. Ss rated their attraction toward five strangers who were .00, .25, .50, .75, or 1.00 similax in attitudes. Half of the Ss rated perceived similarity to the strangers prior to making the attraction ratings (similarity]then liking condition) and the other half rated attraction prior to similarity ratings (liking/then similarity condition). Balance theory would predict an interaction between order of rating conditions and proportion of similar attitudes in determining attraction. Results showed that the only significant effect was due to proportion of similar attitudes. Factor analyses of the rating data showed that attraction and similarity ratings loaded on different factors. This result indicates that perceived similarity does not necessarily mediate attraction. Trend analyses suggested that the reverse may hold to some extent; perception of attraction may partially determine perceived similarity. This result suggests that a restricted balance model may be viable. However, the preponderance of the evidence supported reinforcement theory.     

Revisiting the limits of language: The odor lexicon of Maniq

It is widely believed that human languages cannot encode odors. While this is true for English, and other related languages, data from some non-Western languages challenge this view. Maniq, a language spoken by a small population of nomadic hunter–gatherers in southern Thailand, is such a language. It has a lexicon of over a dozen terms dedicated to smell. We examined the semantics of these smell terms in 3 experiments (exemplar listing, similarity judgment and off-line rating). The exemplar listing task confirmed that Maniq smell terms have complex meanings encoding smell qualities. Analyses of the similarity data revealed that the odor lexicon is coherently structured by two dimensions. The underlying dimensions are pleasantness and dangerousness, as verified by the off-line rating study. Ethnographic data illustrate that smell terms have detailed semantics tapping into broader cultural constructs. Contrary to the widespread view that languages cannot encode odors, the Maniq data show odor can be a coherent semantic domain, thus shedding new light on the limits of language.     

类似性结构组合的实验研究

在国外研究使用点击映射(one—shot mapping)的基础上,本研究在一个实验中从新的角度使用了点击映射的方法探讨了关于类似性的一种观点：类似性判断与差异性判断不是一种对立的镜像关系;研究结果表明。类似性判断与差异性判断都是结构组合。     

The Reduced and Enhanced Impact of Shared Features on Individual Brand Evaluations

In 3 studies, consumers received information about products that had either positive or negative features in common. When they evaluated the products shortly afterward, the common features had little influence on their evaluations, and this was true regardless of whether they had made a preference rating, a similarity rating, or no rating at all before reporting these evaluations. In all cases, consumers apparently used one of the brands spontaneously as a standard of comparison for the other and, in doing so, discounted their common features. When their evaluation was delayed, however, shared features had greater impact. Two factors appear to be operating in the latter conditions. First, the alternative brand is less accessible and, therefore, is less likely to be used as a standard of comparison. Second, shared features are easier to recall than unique features, so they have a disproportional influence for this reason.     

The organization of verbs of knowing: Evidence for cultural commonality and variation in theory of mind

Cross-cultural commonality and variation in folk theories of knowing were studied by examining the organization of verbs of knowing in German and Japanese adults. German and Japanese adults performed one of two tasks: a similarity judgment task and an attribute rating task. Organizational structure was assessed for the similarity judgment task using multidimensional scaling and additive similarity tree analyses. The attribute rating task was used to describe the characteristics that organized the dimensions and clusters emerging from the scaling solutions. The folk theory of mind displayed was an information processing model with constructive components, although the constructive aspects were more salient for the Germans than for the Japanese.     

人格相似性与关系质量的关系：单向视角

温州两所大学的199个大学生在人际形容词量表上对自己的人际特质进行评定,并对自己选定的一个关系对象进行评定,感知与选定对象的人格相似性,同时评定与该对象的关系质量。根据自我报告和评定对象计算出假定相似性。对关系质量的分层回归分析得到人际特质两个维度的交互效应均为正,即表现出假定相似相吸的效应趋势,亲和性维度显著;相关显示假定的相似性和感知的相似性与关系质量均呈中等程度相关;路径分析表明感知的相似性中介假定相似性与关系质量的关系。     

Structural representations of person perception A comparison between own and provided constructs

Abstract.— The dimensions of perceiving other people were compared from two types of data, one representing a person's individual constructs and indexed by two different rating methods (Reptest technique and Similarity Ratings), the other constructs provided by the experimenter, also measured by two methods (Semantic Differential and Personality Differential). Thirty-one female psychology students, rated fifteen role figures by using the four methods. Indices of factorial and cluster similarity of roles showed that the different rating methods yielded highly similar role structures. Each congruent factor and cluster could also be characterized by identical trait properties, but there were also important differences between the methods. Especially family roles formed factors and clusters that were not similar in the four sets of data. The results are discussed within the context of Kelly's personal construct theory.     

A similarity-based range—frequency model for two-category rating data

This research examines whether an extension of range-frequency theory (Parducci, 1965) could successfully predict context effects on discrimination between stimuli for two-category rating data. The new model proposes that similarity comparisons between stimulus values and rating category prototypes take place on an implicit range—frequency scale, and thus it entails effects of context on both mean ratings and discrimination indices. Predictions of the similarity-based range—frequency model were tested in an experiment in which the frequency distribution was manipulated by varying the range and relative frequencies of squares, with participants rating square sizes on a two-category scale. The model provided a parsimonious account of how discriminability of common stimuli increases with a decrease in stimulus range or an increase in the density of stimuli within a subrange.     

An efficient method for obtaining similarity data

Measurements of similarity have typically been obtained through the use of rating, sorting, and perceptual confusion tasks. In the present paper, a new method for measuring similarity is described, in which subjects rearrange items so that their proximity on a computer screen is proportional to their similarity. This method provides very efficient data collection. If a display hasn objects, then, after subjects have rearranged the objects (requiring slightly more thann movements),n(n-1)/2 pairwise similarities can be recorded. As long as the constraints imposed by two-dimensional space are not too different from those intrinsic to psychological similarity, the technique appears to offer an efficient, user-friendly, and intuitive process for measuring psychological similarity.     

Role of Behavior Level,Behavioral Variability,and Ratee Order in the Formation of Appraisal Ratings

Two laboratory studies were conducted to assess the role of behavioral data in the formation of appraisal ratings. In the first study, which included 200 participants, significant main effects using a split-plot factorial analysis of variance design were found for performance variability, performance level, and ratee order. A central drift phenomenon also was discovered with ratees whose performance profiles were both similar and extreme (extremely high or extremely low), receiving ratings that drifted toward average as a function of rating queue position. A second study, which included 126 participants, replicated this drift phenomenon and isolated the cause as profile similarity rather than profile extremity. Implications of these findings for rating accuracy and future "truescore" research designs are discussed.     

Associations between rating content and self-other agreement in multi-source feedback

In a study of the ratings of managers' behaviour made by themselves, their supervisors, subordinates, and peers, it was found that degree of similarity between self-ratings and ratings by colleagues was significantly associated with three aspects of rating content: the observability of a behaviour, its desirability, and the degree to which it concerned technical rather than interpersonal activities. However, these associations with rating content depended on which of two indices of between-rater similarity was studied, in terms of either the correlation or the difference between ratings. The desirability of behaviour more influenced correlation measures of similarity, and the type of behaviour had more influence on difference scores: managers over-rated their interpersonal activities but not their technical effectiveness.     

Alternatives to Difference Scores as Dependent Variables in the Study of Congruence in Organizational Research

In organizational research, difference scores and profile similarity indices are often used as dependent variables in studies predicting the congruence (i.e., fit, match, similarity, agreement) between two constructs. Unfortunately, this practice introduces serious conceptual and methodological problems that render results ambiguous and potentially misleading. This article proposes an alternative procedure that avoids these problems but fully captures the effects of one or more independent variables on the congruence between two dependent variables. This procedure is illustrated by reanalyzing data from a study of feedback seeking and rating accuracy (Ashford & Tsui, 1991), and the results of this study are reinterpreted in light of these analyses. Limitations and areas for further development of the procedure are discussed.     

Effects of categorical dissimilarity and affective similarity between constituent words on metaphor appreciation

This study examines the effects of categorical and affective similarities between the constituent words of sentence (Noun-A is like Noun-B) on metaphor appreciation. In the preliminary studies using 48 words to be used later, the two kinds of similarity were measured separately by either sorting or semantic differential technique. In the main study, a set of five rating tasks concerning three types of formulaic sentences (literal, metaphorical, and anomalous) was given for each of 28 subjects in relation to (a) similarity between the words in a pairing, (b) similarity of the words in a particular formulaic sentence, (c) sentence comprehensibility, (d) sentence novelty, and (e) metaphor aptness. The application of path analysis for the above data revealed the following: (1) Categorical dissimilarity affects novelty [N], (2) affective similarity influences comprehensibility [C], and (3) [N] and [C] affect metaphor aptness.A portion of the results was reported at the First Annual Conference of the Japanese Cognitive Science Society, Kyoto, July 1984.     

Sources of interference in the attentional blink: target-distractor similarity revisited

Observers monitored streams of words or letters (10 items/sec) for one or two targets. An attentional blink (AB) effect was observed in which identification of the first target temporarily impaired identification of the second target. Target identification was impaired when the distractors were composed of either letters or false-font characters (cf. Maki, Couture, Frigen, & Lien, 1997). An asymmetrical AB effect was observed with letters and mathematical symbols; the AB effect was largest for symbol targets and letter distractors. The characters used in these experiments were rated on their meaningfulness, familiarity, and other stimulus properties. The rating data showed that pixel density best accounted for the asymmetrical target-distractor similarity effects. Modulation of the AB effect by target-distractor similarity appears to result partly from low-level masking. But masking effects may be reduced by attentional capture by target features.     

The accuracy of roommate ratings of behaviors versus beliefs

We evaluated the accuracy of peer ratings of roommates’ personality characteristics, against roommate self-ratings, as a function of rating domain observability. Instead of the usual ratings of broad personality traits, however, our domains represented peer ratings of narrow exemplars of personality traits. Specifically, we compared roommate ratings on (a) observable trait-related behaviors with (b) unobservable trait-related attitudes or beliefs. We observed greater self-peer agreement in rating behaviors, in general, than in rating beliefs. We also observed greater tendency of raters to adopt an assumed similarity heuristic when judging their roommates’ attitudes and beliefs than their behaviors. We discuss the contribution of these findings to understanding the determinants of accuracy in personality judgments and developing best practices for personality assessment.