An investigation of the relationship between category development and performance on a standardized test of reading

Background knowledge has long been considered to be central to reading. Categories of information may be one way of viewing the organization of background knowledge. Historically, categories have been considered to be fixed entities, developing only as additional similar exemplars are added. Recent research, however, challenged this. Five non-quantitative components or classifications of categories were identified in the literature. These were: (1) defining features, (2) prototypes, (3) natural kinds, (4) non-basic level categories, (5) taxonomies. It followed that reading development might be associated with category development of these.

Five questions, then, based on the relationship between each developmental component of categorization and both reading level and grade level were asked. These questions required measurement tools for reading development and categorization development.

The Gates-MacGinitie Reading Tests (1989) Level 2, Form K was used to measure reading development. The test was used to identify low readers (stanines one, two, and three) and high readers (stanines seven, eight, and nine) in first and third grade children in two rural east Tennessee elementary schools.

The literature supported the use of sorting tasks as measures of categorization. Therefore, five sorting tasks were used to evaluate the categorizations of first and third grade low readers and high readers.

The following results indicated a pattern of development: 1. Moderate to strong associations between reading stanine levels and many of the categorization tasks for both grade levels. 2. A pattern of differences between the means of low readers and high readers. 3. A comparison of first grade low readers and third grade high readers which suggested that the effects of both reading level and grade level increased the magnitude of the group differences.

The results suggested that category development could associated with reading development. This has be implications for reading evaluation and instruction. Further research is needed for verification.