Patterns of development in conceptual styles and word association



Journal Title

Journal ISSN

Volume Title



The objectives of this study were: (a) To investigate the patterns of development in children's behavior in conceptualisation and word association; (b) to determine whether or not the two forms of the "Conceptual Styles Test" were comparable; and (c) to explore the relationships between certain types of performance across tasks. The manner in which children's thinking progresses toward an adult level was of particular interest. Some disagreement exists as to whether cognition progresses through a series of discrete stages, each specific to a given age range, or whether it develops in a more gradual, more or less linear fashion as a function of age. The measures used in this study were all "type of operation" measures; that is, they were of a class postulated by Werner (1957) to reflect "discontinuity" in cognitive development. The "Conceptual Styles" tests used in this study were those developed by Kagan and the Fels Research Institute staff (CST) and by Sigel (CST-A). Both require the subject to choose two of a group of three pictures on the basis of conceptual similarity. Responses were scored by the method developed by Kagan, Moss, and Sigel (1960, 1963). According to this system the response is classified as (a) "relational" when it involves a functional relationship between the two chosen stimuli; (b) "analytic" when it involves similarity of objective aspects of the two stimuli; and (c) "categorical-inferential" when it is based on an inferred characteristic or conforms to language convention. The word association test used in this study consisted of 33 homonyms, homographs, and/or words that could be used as more than one part of speech. Responses were classified according to "idlodynamic sets" described by Moran, Mefferd, and Kimble (1964). These sets involve (a) functional responses, (b) contrast and coordinate responses, and (c) synonym and superordinate responses. In addition, Thorndike-Lorge frequency of usage scores and two commonality scores, one derived from the present sample and one from an adult sample, were used. A "preconceptual" classification was employed for both the Conceptual Styles and word association tests for scoring responses that failed to satisfy the criteria of any of the three categories in each scoring system. The 80 subjects used in this study were divided into four age groups of ten males and ten females each. The ages of the groups were six, eight, ten, and twelve. The eleven measures were first analyzed for age changes. All measures except functional word association responses varied significantly with age. The relational and categorical-inferential scores from the CST-A and the synonym-superordinate, adult commonality scores from the word association test showed only significant linear trends, reflecting a steady rate of increase or decrease across all age groups. Both significant linear and cubic trends were found for the preconceptual measure from the CST-A; mean scores for this measure showed a change in direction at age ten--an increase-only to resume a decrease among twelve year olds. Scores for the analytic CST-A measure, and the contrast-coordinate, peer commonality, Thorndike-Lorge value, and preconceptual word association measures showed both significant linear and quadratic trends as a function of age. These results indicated that, for these five measures, there was a relatively "discontinuous" growth pattern, because of a significant change in the rate of change at some point along the age curve. Comparisons between the two Conceptual Styles tests, and between the CST-A, developed by Sigel, and the word association test were made by use of Pearson product moment correlations. The relationship between the two Conceptual Styles tests was generally significant across all scoring categories only among the eight year old group. The generally insignificant correlations found in the other age groups appeared to be largely attributable to differences in construction of the tests. The CST was developed to elicit relational and/or analytic concepts; and, insofar as possible, to suppress categorical-inferential concepts. The CST-A was not so constructed. Correlations between the more general of the Conceptual Styles tests (CST-A) and word association measures all failed to meet an acceptable level of significance. In all comparisons, except those between analytic and contrast-coordinate responses, such relationship as was found showed a change in direction. Intercorrelations between each of the two commonality word association measures and Thorndike-Lorge values were generally significant across all age groups. The two exceptions were those correlations involving the ten year old group. Both of these relationships were at about a zero level. This finding suggests that some sort of transition might be occurring at this age which may be partly attributable to changes in language usage patterns. Consideration of qualitative differences occurring within the various categories on both the Conceptual Styles test and the word association test suggested that there was probably not sufficient homogeneity within the scoring categories of either test to provide a definitive test of "true" relationship between perceptual-conceptual and verbal development, without further refinement of the scoring procedures. Suggestions for future research involving the Conceptual Styles tests included analysis of differences occurring within the various categories at different ages, investigation of the possible relationship between the various categories used and the nature of the stimulus array, and consideration of the "regressive" aspect apparent in some of the responses. Some difficulties arising both from the scoring system and from the specific words used on the word association test were described. No analysis of these data was done on the basis of the sex of the subjects, although sex differences were apparent. This area was suggested as a fruitful one for future research on both tests.



Association tests, Concepts, Child development