Conceptual styles: Developmental patterns and consistency across tasks

Date

1971

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The two main purposes of the present study were (1) to investigate the patterns of development in children’s sorting behavior in terms of the 'continuity� or 'discontinuity' with which certain types of behavior change, and (2) to investigate the consistency of such behavior across tasks of varying content. The three tests used were an object sorting test, a conceptual styles test (SCSI), and the Role Construct Repertory Test (RCRT). Two additional purposes involved (a) investigating the relationship between scores obtained with the active and passive phases of the object sorting test and, (b) investigating possible relationships between the nature of test items and the type of response given on SCSI. Two scoring systems, the McGaughran conceptual area analysis and the conceptual styles measures, as modified by Sigel (1967), were applied to all tests. The McGaughran conceptual area measures are based on two dimensions. The first dimension, open- closed, refers to the relative highness or lowness of order of conceptual classification, while the second dimension, public-private, reflects the relative communicability of the concept. Both dimensions are scored simultaneously resulting in four types of derived measures of conceptual responses: Open-public, closed-public, closed-private, and open-private. The conceptual styles measures represent a more perceptually-oriented system of analysis. Two classes of responses, descriptive-global and descriptive part-whole, are distinguished on the basis of differentiation of the perceptual cues used. A third class, categorical-inferential, is based on relative conformance to language convention, while the fourth class, relational-contextual, is a contextbound type of concept. An additional measure, cognitive complexity, was derived from RCRT according to Bieri’s original method (Blerl, 1955). In all 39 measures were used. The eighty subjects used in this study were divided Into four age groups of ten boys and ten girls each. The ages of the groups were 7, 9, 11, and 13 years. Analyses of variance were first computed to determine whether each of the various measures changed significantly as a function of age. Those 25 measures showing significant changes with age were then analyzed for trends. The developmental patterns were then classified according to the three criteria—predominance, directionality, and rate of change—developed by Penk (1965). The McGaughran conceptual area measures generally showed a more 'continuous,' linear change with age, particularly when derived from the object sorting test, than did the conceptual styles measures. The conceptual styles measures were more 'discontinuous' In that they more often showed changes In directionality and In their rate of change across age groups. Pearson product moment correlations were then computed between all measures bearing the same name across ages and tests, except In comparisons of the McGaughran conceptual area measures with similar measures derived from the RCRT. For those comparisons, Spearman rank order correlations were used. Examination of correlation matrices indicated that, when results for age groups were combined, a fair degree of relationship existed between operations bearing the same names for three of the four McGaughran conceptual area measures in all comparisons involving the object sorting test and SCSI, When correlations were examined for separate age groups, the patterns of significant relationships were sparse and appeared to be differentially affected by age. This was particularly true in comparisons involving the active and passive phases of the object sorting test where no significant relationships were found among the two younger groups. The closed-public measure from this system showed the least evidence of consistency across tests. A fair degree of relationship was also found for the conceptual styles scoring system across tests when results for age groups were combined, but again the relationships were less frequent when the age groups were considered separately. The patterns of correlations across tests indicated that the different types of response classes showed varying degrees of relationships. The extent of consistency appeared to be affected both by age and by the nature and content of the task. Some correspondence between the two systems was found. This correspondence was greatest for the two most divergent measures with in each system. Some overlapping was found, in which a measure from one system was related to several measures from the other system. The degree of correspondence appeared to be affected both by the nature of the test and by the presence of technique factors. Significant correlations were found between scores obtained on the two phases of the object sorting test for three of the four McGaughran conceptual area measures and for two of the four conceptual styles measures. Some support was obtained to indicate that the type of material being classified in separate SCSI items influenced the nature of the conceptual response; this relationship also appeared to be differentially affected by age. The relationships between tests were also considered in terms of the amount of constraint imposed by test instructions, relative difficulty, and perceptual versus conceptual content insofar as such differences affected both developmental patterns and the amount of consistency found. Suggestions for future research included some refinement of particular measures within each of the two systems, as well as examination of sex differences and their effects upon the developmental patterns and nature and degree of consistency found.

Description

Keywords

Child development, Cognition in children

Citation