Abstract
Causal priorities between word meaning ability and paragraph meaning ability were examined in three types of readers across a period of nine months. Subjects were 180 pupils in grades 7 and 8. They were categorized by relative performance in word and paragraph meaning into three groups (n = 60 each) defined as “balanced”, “word dominant” and “paragraph dominant”. Cross-lagged correlation analysis supported the causal priority of word meaning in balanced and word dominant readers and a different pattern in the paragraph dominant readers. For most readers, the conventional wisdom hypothesis of word meanings as building blocks of paragraph meaning received empirical support.
