Ordinal size scaling in preschool children.

Abstract

Young children are limited in their usage of comparative adjectives and ordinal numbers, typical ways of describing ordinal relationships. However, research in a number of areas suggests the possibility of a precursor level of ordinal concept. To facilitate the search for precursor ordinal skills, ordinal ability was defined in ordinal measurement terms. Only "greater than - less than," asymmetric judgements were required. Additionally, linguistic demands were reduced by using family-role terms as size designators. Experimental manipulations included variations in scale size and in the complexity level of ordinal conceptualization. Solution strategies based on "good form" and on "pairwise comparison" were precluded by using pictures of randomly placed objects which could not be manipulated by the child. Ninety-six 3-6 year old children pointed to "Daddy," "Mommy," "Big boy/girl," "Little boy/girl," and "Baby" when shown sets of 3 to 5 circles or squares which differed only in size. Tasks were of three types: Identification, mapping labels onto a single set of objects; Coordination, mapping labels onto two identical sets of objects in which corresponding "family members" are the same size; and Transposition, mapping labels onto two separate sets in which corresponding family members are of different sizes. Data were analyzed in an Age (3), by Scale Size (3), by Complexity Level (3), by Shape (2) mixed design ANOVA, and significant main effects were obtained for all variables. Tasks became more difficult with increases in scale size, and in complexity level. Square objects were slightly more difficult than circular, and older children were more proficient than younger ones. Post hoc tests generally supported the obtained main effects. Finer grained analysis using Latent Trait procedures supported the global ANOVA results, and supported the hypothesis that the end points of a scale are easier than the central positions. Response patterns indicated that errors were size-related, and suggested transitional levels of performance. The present study demonstrates that children as young as three can demonstrate a precursor ordinal concept when the task is framed in familiar terms and is placed in a context which is meaningful for them.