ERIC Number: ED421674
Record Type: RIE
Publication Date: 1998-Jul
Reference Count: N/A
Category Induction and Hierarchical Classification Assessed by Property Inference: The Influence of Complexity.
Halford, Graeme S.; Andrews, Glenda; Jensen, Ingalise
Relational complexity has been found to be an effective metric for cognitive tasks. The hypothesis that the greater difficulty and later age of attainment of hierarchical classification as compared to category induction are attributable to differences in structural complexity was tested. Hierarchical classification entails a ternary relation between Categories B, A, and A prime such that A and A prime are included in B. Category induction entails a binary relation between a category and its complement. Forty children, 3 to 6 years old, were assessed on hierarchical classification by property inference between levels (basic-subordinate or subordinate-basic) and on category induction using property inference within levels (basic-basic, subordinate-subordinate). The same hierarchies were used in both tasks, and special care was taken to control for question content. As predicted from relational complexity theory, hierarchical classification was more difficult than category induction, and children older than 5 years succeeded on both tasks, but 3-year-olds succeeded on category induction only. Multiple regression analysis showed that 68% of the age-related variance was accounted for by performance on tasks from other domains that were known to entail the same level of relational complexity. (EMK)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A
Note: Poster presented at the Biennial Meetings of the International Society for the Study of Behavioural Development (15th, Berne, Switzerland, July 1-4, 1998).