Developmental trends in free classification: Evidence for a new conceptualization of perceptual development

Two studies are reported to explore the hypothesis that young children perceive integrally some stimuli that older children perceive separably. In both, kinder-garteners, second graders, and fifth graders (approximately 5, 8, and 11 years old) are required to classify sets of stimuli that vary in size and brightness. Triads are used in Experiment 1 and tetrads are used in Experiment 2. Also, in Experiment 2, second classifications, judgments of which classification is “best,” and verbal justifications for classifications are obtained. The general finding is that the kinder-garten data systematically implicate integrality of size and brightness while the fifth-grade data systematically implicate separability of size and brightness. The second-grade data are more ambiguous. Issues related to refining the developmental hypothesis and to extending its supportive data base are considered in a final discussion.     

Importance of the overall similarity of objects of adults' and children's classifications

Previous work has shown that adults and older children tend to classify multi-dimensional objects by identity on one dimension, whereas children under 8 years of age tend to classify these same objects by a relation of overall similarity. The present study investigated the hypothesis that this developmental trend is restricted to the classification of simple objects that differ only by limited amounts on a few dimensions. The specific hypothesis was that overall-similarity relations structure both adults' and children's classifications of heterogenous objects (objects that differ in a variety of ways). This hypothesis was suggested by the correspondence between the structure of young children's classifications and the structure of natural categories. The result of two experiments supported the hypothesis. When the to-be-classified objects varied simultaneously on relatively many dimensions, adults as well as children constructed classifications that maximized within-category similarity on all varying dimensions. The implications of these results for accounts of the perception of multidimensional relations and classificatory behavior are discussed.     

Is there a developmental trend from integrality to separability in perception?

Some previous literature suggests that young children perceive in an integral, holistic fashion stimuli that older children perceive in a separable, dimensionalized mode. A prediction from a strong form of this position is that younger children actually may perform more rapidly a speeded classification task that requires “condensation” than a task that requires “filtering” (if the similarity relations among the wholes favor the former task). Older children should be able to take advantage of the simple unidimensional basis of the filtering task and thus accomplish it much more rapidly than the condensation task. The results are only partially in accord with the predictions. Kindergarteners (5 years of age), on size-and-brightness stimuli, show no speed advantage on either task, while second (8 years) and fifth (11 years) graders clearly show more rapid filtering. Therefore, the developmental hypothesis is in need of some revision and elaboration. Some stimuli are less separable for younger than for older children, but even five year olds can access their dimensional structure under some conditions.     

Perceptual sensitivity to dimensional and similarity relations in free and rule-governed classifications

Two experiments were conducted to investigate the perceptual primacy of dimensional and similarity relations in the stimulus classifications of younger and older subjects. In Experiment 1, 4- and 10-year-olds were given free classifications in which they could group stimuli according to overall similarity or identities in size, color, or orientation. Both age groups classified stimuli most frequently according to identities on separate dimensions. In Experiment 2, 4-year-olds and adults were given free classifications followed by rule-governed classifications which required them to group stimuli according to specific relations. In the free classifications, a majority of subjects in both age groups classified the stimuli most frequently according to identities on separate dimensions. In the rule-governed classifications, both age groups were more accurate when a single separate relation was required for solution than when overall similarity was required. These results support a differential-sensitivity view of perceptual development, which asserts that individuals at all ages primarily perceive and use separate relations.     

Exploring and reexploring issues of integrality,perceptual sensitivity,and dimensional salience

Some observations that can be described conveniently by investigators in terms of dimensional sensitivity and dimensional salience actually need not implicate the psychological reality of the dimensions for the subjects. The developmental hypothesis that stimuli are perceived often as integral in early childhood can account for such phenomena only with the assumption that young children often apprehend objects as global wholes, related to one another by overall similarity. Classification data, recently presented by J. R. Aschkenasy and R. D. Odom (Journal of Experimental Child Psychology, 1982, 34, 435–448) can be reinterpreted in this way. On this reinterpretation, Aschkenasy and Odom's findings are exactly what the developmental form of the integrality-separability hypothesis predicts. Moreover, a variety of other findings and lines of inference converge on the conclusion that, in young children, the use of overall similarity relations to organize perception and cognition predominates over the use of dimensional relations (and that the relative balance between these stimulus relations changes in development). Examples are selected from the literature on the natural development of word meanings, as well as from the literatures on children's and adults' classification.     

Classifying by dimensions and reading: a comparison of the auditory and visual modalities

Children have difficulty learning to read alphabetic writing systems, in part, because they have difficulty segmenting spoken language into phonemes. Young children also have difficulty attending to the individual dimensions of visual objects. Thus, children's early difficulty in reading may be one sign of a general inability to selectively attend to the parts of any perceptual wholes. To explore this notion, children in kindergarten through fourth grade (Experiments 1, 3, and 4) and adults (Experiment 2) classified triads of spoken syllables and triads of visual objects. Classifying speech by common parts was positively related to reading and spelling ability (Experiments 1 and 4), but usually not to classifying visual stimuli by common parts under free classification instructions (Experiments 1 through 3). However, classification was more consistent across the visual and auditory modalities when the children were told to classify based on a shared constituent (Experiment 4). Regardless of instructions, performance on the visual tasks did not usually relate to reading and spelling skill. The ability to attend selectively to phonemes seems to be a "special" skill--one which may require specific experiences with language, such as those involved in learning to read an alphabetic writing system.     

Phonological deficits: Beneath the surface of reading-acquisition problems

Summary In Experiment 1, the performance of young retarded readers on speech-segmentation tasks was compared with the performance of normal subjects matched for chronological age (CA) and with subjects matched for reading age (RA). Retarded readers were poorer than both control groups in consonant deletion, while there was no difference between the groups on a rhyme-judgement task and a syllabic-vowel-reproduction task. In Experiment 2, another group of reading retarded children was compared with CA and RA controls on the classification of pseudowords, either by common phoneme or by overall phonetic similarity. The retarded readers made fewer classifications based on common phoneme than both control groups, while there was no difference between the groups in classifications based on overall phonetic similarity. In Experiment 3, adult developmental dyslexics were compared with normal adults in the tasks of Experiments 1 and 2. The dyslexics made fewer classifications based on a common phoneme than the normals, while no difference was found in classifications based on overall phonetic similarity.     

Classification behavior and measures of intelligence: dimensional identity versus overall similarity

Individuals tend to adopt either analytic or holistic modes of categorizing objects. In two studies, we examined the relation between these categorization tendencies and cognitive abilities as measured by standard psychometric instruments. The participants in both studies were pretested with a restricted classification task in which it was possible for them to classify simple stimuli by dimensional identity or overall similarity. Those making a large number of either type of categorization were then tested with subtests of the WAIS-R and with the Raven's progressive matrices. Across both studies, the analytic individuals (many dimensional identity classifications) scored higher than the holistic individuals (many overall similarity classifications) on some but not all of the subtests. The results are consistent with the idea that holistic modes of categorization may be more "primitive" than analytic modes. The findings are discussed in terms of the association between categorization mode and either general or specific cognitive abilities.     

Classification training in two- and three-year-old children

Two- and three-year-olds who did not group a set of geometric stimuli according to complete similarity on a pretest were divided into three experimental conditions: (a) a modeling condition in which the subjects observed a model group according to complete similarity while verbalizing his strategy, (b) a reinforcement condition in which the subjects were reinforced for sorting similar stimuli into boxes, and (c) a control condition. After the training sessions the Ss were given two posttests—one with stimuli identical to those used during the training session and one with a different set of geometric stimuli. On the posttests significantly more similarity classifications were obtained in the modeling condition than in either the reinforcement or the control conditions, which did not differ. It was concluded (a) that two- and three-year-olds can learn to group according to similarity, and (b) that modeling with verbalization is a very effective method for teaching young children to classify according to similarity.     

Classifying multidimensional stimuli: stimulus, task, and observer factors

When observers decide how to classify stimuli, they often employ one of two types of information: identity along one particular dimension or overall similarity. The present studies examined interrelations among the factors which determine the use of these types of information. Participants' classifications of certain types of materials (e.g., size and brightness, length and density) revealed strong individual differences, were related to the individual's response tempo and selective processing ability, and were influenced by task demands. Classifications of other materials (e.g., saturation and brightness) did not reveal individual differences, were not affected by response tempo and selective processing ability, and were unaffected by changes in task demands. The former, but not the latter, types of materials have also been found to be influenced by developmental differences. The results are consistent with the idea that differences in response tempo and selective processing ability underlie observer differences (both individual and developmental) and that certain types of stimuli which are not susceptible to such influences set boundary conditions for observer differences. The results are discussed within an integral-to-separable model of processing.     

Perception of multidimensional stimuli: a differential-sensitivity account of cognitive processing and development.

A differential-sensitivity account of cognitive processing is described that emphasizes the development of perceptual sensitivity to object relations that are directly perceived. Four experiments are presented that investigate this account and compare it to the integrality-separability view of development and the view that younger children are nonselective in their processing of multidimensional stimuli. Results show that stimulus differences are more salient than identities (Expt. 1), younger as well as older children are highly selective in their perception and classify stimuli by separate dimensions (Expt. 2), differential sensitivity affects the perceived magnitude of stimulus differences (Expt. 3), and younger and older children perceive separate dimensions in speeded classifications (Expt. 4). The importance of considering individual patterns of responses in cognitive and developmental research is also demonstrated.     

Three aspects of cognitive development

An attempt was made to characterize and explain developmental differences in children's thinking, specifically in their understanding of balance scale problems. Such differences were sought in three domains: existing knowledge about the problems, ability to acquire new information about them, and process-level differences underlying developmental changes in the first two areas. In Experiment 1, four models of rules that might govern children's performance on balance scale problems were proposed. The rules proved to accurately describe individual performance and also to accurately predict developmental trends on different types of balance scale problems. Experiment 2 examined responsiveness to experience; it was found that older and younger children, equated for initial performance on balance scale problems, derived different benefits from identical experience. Experiment 3 examined a potential cause of this discrepancy, that younger children might be less able than older ones to benefit from experience because their encoding of stimuli was less adequate. Independent assessment procedures revealed that the predicted differences in older and younger children's encoding were present; it was also found that these differences were not artifactual and that reducing them also reduced the previously observed differences in responsiveness to experience. It was concluded, therefore, that the encoding hypothesis explained a large part of the developmental difference in ability to acquire new information.     

A purely power-space association without spatial and strategic biases

Previous studies have shown that powerful and powerless concepts are metaphorically associated with top and bottom spaces respectively. However, this association might be contaminated by spatial and strategic biases due to the involvement of spatialized stimuli or responses. It is unknown whether power by itself can automatically activate spatial representations. To eliminate spatial and strategic biases, Experiment 1 separately presented power and spatial stimuli at the center of the screen, and participants had to classify power words and HIGH/LOW labels (Experiment 1a) or indicate up/down arrows (Experiment 1b) using a single key and follow different response rules that were combined with a Go/no-go task. Experiment 2 replicated the non-spatialized design but used an implicit power judgment task. Based upon the non-spatialized method, the results provide the first evidence that a power-space association effect still exists when eliminating spatial and strategic biases, revealing the nonconscious spatial nature of power.     

Lexical restructuring in the absence of literacy

Vocabulary growth was suggested to prompt the implementation of increasingly finer-grained lexical representations of spoken words in children (e.g., [Metsala, J. L., & Walley, A. C. (1998). Spoken vocabulary growth and the segmental restructuring of lexical representations: precursors to phonemic awareness and early reading ability. In J. L. Metsala & L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 89-120). Hillsdale, NJ: Erlbaum.]). Although literacy was not explicitly mentioned in this lexical restructuring hypothesis, the process of learning to read and spell might also have a significant impact on the specification of lexical representations (e.g., [Carroll, J. M., & Snowling, M. J. (2001). The effects of global similarity between stimuli on children’s judgments of rime and alliteration. Applied Psycholinguistics, 22, 327-342.]; [Goswami, U. (2000). Phonological representations, reading development and dyslexia: Towards a cross-linguistic theoretical framework. Dyslexia, 6, 133-151.]). This is what we checked in the present study. We manipulated word frequency and neighborhood density in a gating task (Experiment 1) and a word-identification-in-noise task (Experiment 2) presented to Portuguese literate and illiterate adults. Ex-illiterates were also tested in Experiment 2 in order to disentangle the effects of vocabulary size and literacy. There was an interaction between word frequency and neighborhood density, which was similar in the three groups. These did not differ even for the words that are supposed to undergo lexical restructuring the latest (low frequency words from sparse neighborhoods). Thus, segmental lexical representations seem to develop independently of literacy. While segmental restructuring is not affected by literacy, it constrains the development of phoneme awareness as shown by the fact that, in Experiment 3, neighborhood density modulated the phoneme deletion performance of both illiterates and ex-illiterates.     

Verbal rehearsal and selective attention in children with learning disabilities: A developmental lag

To investigate the development of verbal rehearsal strategies and selective attention in learning disabled children, Hagen's Central-Incidental task was administered to younger learning disabled (M CA = 8.68 years) and normal (M CA = 8.62 years) boys in Experiment 1 and to intermediate (M CA = 10.18 years) and older (M CA = 13.48 years) learning disabled boys in Experiment 2. Also, in Experiment 2, an experimentally induced verbal rehearsal condition was included to determine its effects on serial recall and selective attention performance. In Experiment 1, the serial postion curve of the normals revealed both a primacy and a recency effect, whereas that of the learning disabled revealed a recency effect only. In Experiment 2, both the intermediate and the older learning disabled exhibited both primacy and recency effects under both standard and rehearsal conditions. A developmental analysis of central recall for the three learning disabled groups revealed constant age-related increases in overall central recall and in primacy recall. That the normals recalled more central, but not more incidental, information than the learning disabled in Experiment 1 suggests that the learning disabled are deficient in selective attention. Correlational findings suggest that the selective attention of the learning disabled improves with age. The results were interpreted as support for the hypothesis of a developmental lag in the learning disabled population.     

A computational account of the development of the generalization of shape information

Abecassis, Sera, Yonas, and Schwade (2001) showed that young children represent shapes more metrically, and perhaps more holistically, than do older children and adults. How does a child transition from representing objects and events as undifferentiated wholes to representing them explicitly in terms of their attributes? According to RBC (Recognition‐by‐Components theory; Biederman, 1987 ), objects are represented as collections of categorical geometric parts (“geons”) in particular categorical spatial relations. We propose that the transition from holistic to more categorical visual shape processing is a function of the development of geon‐like representations via a process of progressive intersection discovery. We present an account of this transition in terms of DORA ( Doumas, Hummel, & Sandhofer, 2008 ), a model of the discovery of relational concepts. We demonstrate that DORA can learn representations of single geons by comparing objects composed of multiple geons. In addition, as DORA is learning it follows the same performance trajectory as children, originally generalizing shape more metrically/holistically and eventually generalizing categorically.     

Is overall similarity classification less effortful than single-dimension classification?

It is sometimes argued that the implementation of an overall similarity classification is less effortful than the implementation of a single-dimension classification. In the current article, we argue that the evidence securely in support of this view is limited, and report additional evidence in support of the opposite proposition—overall similarity classification is more effortful than single-dimension classification. Using a match-to-standards procedure, Experiments 1A, 1B and 2 demonstrate that concurrent load reduces the prevalence of overall similarity classification, and that this effect is robust to changes in the concurrent load task employed, the level of time pressure experienced, and the short-term memory requirements of the classification task. Experiment 3 demonstrates that participants who produced overall similarity classifications from the outset have larger working memory capacities than those who produced single-dimension classifications initially, and Experiment 4 demonstrates that instructions to respond meticulously increase the prevalence of overall similarity classification.     

Development of classification: The use of similarity and dimensional relations

Findings about perceptual development indicate that overall similarity is the primary perceptual relation by which young children compare complex objects. Traditional studies of classification, however, did not focus on children's organizational use of holistic relations but rather on their ability to classify by dimensions or criterial attributes. The results from such traditional studies suggest that young children are deficient classifiers. The present research investigated the possibility, contrary to the traditional view, that 4- to 6-year-old children are competent and systematic classifiers at least by overall similarity. In three experiments, preschoolers and kindergarteners classified various sets of multidimensional stimuli that could be organized into categories by overall similarity or by dimensional attributes. Consistent with the research in perceptual development, the children were highly attentive to overall similarity. However, the preschoolers in particular showed marked difficulty in using this relation to form categories of more than two objects. The children's difficulties were highly reminiscent of traditional claims about early classification. Analyses of the classification strategies used by the children, however, suggest that even the youngest children understood the purpose of a classification. The developmentla changes appear to be in the ability to execute a classification. Importantly, type of classification strategy was independent of type of category organization. Individual children used the same strategies both when classifying by overall similarity and by dimensional attributes. These results strongly suggest that it is the classification skills themselves, and not just the ability to classify by particular relations, that change with age.     

The generality of transfer: Effects of age and similarity of training and transfer tasks

Two experiments were conducted investigating the extent to which transfer of training would take place. The Ss were trained to attend to a particular dimension (color or form) of a set of stimuli and were then presented with a transfer problem. For half the Ss the transfer problem was an intradimensional (ID) shift, and for the remaining half an extradimensional (ED) shift, with the ID-ED difference then being used as an index of the magnitude of transfer. Previous data had indicated that preschool and kindergarten children did not effectively transfer learned attention responses from one task to another if the formats of the two problems were different. Experiment I investigated the hypothesis that older children would show broader patterns of transfer. However, the patterns obtained with second and fifth graders were extremely similar to those obtained with younger children. In Exp. II, it was found that transfer did take place across different task formats if the stimuli remained the same from training to transfer. The data were discussed in terms of the role of contextual cues, or context markers, in the origical acquisition of information and the subsequent transfer (use) of that information.