The Impact of Word-Based and Situation-Based Knowledge on Reading Speed in Adults and Children: A Self-Paced Moving Window Study

The purpose of this study was to examine the manner in which word-based knowledge and situation-based knowledge affect reading speed in typically developing (TD) 3rd grade children and adults. Research has demonstrated that both word-based and broader situation-based information are critical for optimal text comprehension, but their relative contributions to reading speeds in children and adults is unknown. In the current study, a sample of 36 adults and 36 children from the greater Katy, Texas area silently read sentences using self-paced moving window computer software. Sentences were constructed such that half had main clauses containing agent:action pairs with a relatively high probability of co-occurrence (e.g., . . .the bird flew from the nest) and half had relatively improbable agent:action pairs (e.g., . . .the bird fell from the nest). In addition, main clauses with improbable relationship were preceded by either a dependent clause containing situation-based information that was expected to make the improbable agent:action relationship more likely (e.g., After hurting its wings…), or a neutral dependent clause containing situation-based information that was not expected to affect the improbable relationship (e.g., After flapping its wings…). Three research hypotheses were tested: 1) Adult and child reading times will be faster for probable than for improbable agent:action relationships when preceded by a neutral dependent clause; 2) Adult and child reading times for improbable agent:action relationships will be faster when preceded by a biasing dependent clause than when preceded by a neutral dependent clause, and 3) The situation-based knowledge in the biasing dependent clause will have a relatively greater impact on adults than on children. Results supported the first hypothesis, showing that both adult and child readers were sensitive to the probability of word-based agent:action relationships. By contrast, situation-based information did not significantly affect reading times for improbable agent:action relationships at either age, and there was no significant interaction between age and condition. The current study is the first to use a moving window display to examine the relative effects of these factors on reading speed of adults and children as they process connected text. This study is also the first to carefully control for both syntactic and semantic characteristics, which have mostly been independently manipulated for investigation in prior studies.