Semantic priming and working memory capacity: A test of distinctive roles in reading comprehension

The question of how readers put together semantic information to construct overall meanings of sentences has long been studied without reaching definitive answers. Recently, activation and working memory capacity have been related to reading through theory and empirical evidence. However, there are inconsistencies in the definitions of these memory constructs by different researchers, and the nature and distinctiveness of the role each plays in comprehension are unclear. The current research used a correlational approach to investigate the potential roles of automatic and controlled memory processes in reading comprehension. Measures of target memory processes were obtained and correlated with measures of comprehension. It was hypothesized that individual differences in both automatic and attention-driven memory processes among a sample of sixth-grade students would be related to individual differences in students' reading comprehension skills. It was also hypothesized that the two memory processes would make distinct contributions to the prediction of comprehension skill.The results of this study were supportive of the contention that the semantic priming measure would predict reading comprehension ability, and that the contribution of semantic priming would be distinct from working memory, lexical processing speed, and verbal knowledge. Priming effects among the child subjects were of a moderate magnitude and were statistically reliable. Semantic priming correlated with the reading comprehension variable r =.60 and was the most powerful predictor of the reading comprehension variable after verbal knowledge. Working memory correlated r =.44 with reading comprehension, and lexical processing speed correlated r =.33. Counter to expectations, lexical processing speed did not make a significant contribution to the prediction of reading comprehension ability. These results provide evidence of the distinctiveness of automatic memory processes such as activation and controlled, attention-driven processes. Each appear to play distinct, important roles in specific reading comprehension processes.The sample consisted of 94 sixth-grade students from two elementary schools in a large western city. General reading ability level was measured from Stanford Achievement Test (SAT) scores, teacher ratings, and performance on a Sentence Verification Test (SVT). Each student performed several computerized tasks measuring semantic priming, working memory, lexical processing speed, and reading comprehension ability, in four 30-min sessions held on separate days of the same week. The contributions of semantic priming, working memory, lexical processing speed, and verbal knowledge (as measured by SAT Vocabulary scores) to the prediction of reading comprehension were analyzed in a regression analysis.