Remembering sentences is not all about memory: The relationship between syntax, working memory, and reading comprehension

Remembering sentences is not all about memory: The relationship between syntax, working memory, and reading comprehension

First Author: Mads Poulsen -- University of Copenhagen
Additional authors/chairs: 
Jessie Leigh Nielsen; Rikke Vang Christensen
Keywords: Syntax, Working memory, Grammar and syntax, Reading comprehension, Assessment
Abstract / Summary: 

Purpose: Recent studies have found correlations between measures of reading comprehension and syntactic skills, that is proficiency in using word order signals to establish semantic relations. However, the task demands of syntactic tests are not very well understood. The present study investigated the properties of two measures of syntactic proficiency. The questions were 1) to what extent the measures correlated with each other (convergent validity) vs how well they correlated with decoding, vocabulary, and working memory (discriminant validity), 2) to what extent variance in the syntactic tasks was explained by vocabulary and memory, 3) and to what extent shared or unique variance in syntax predicted reading comprehension?
Method: Eighty-six grade 6 students completed tests of reading comprehension, decoding, vocabulary, working memory, and syntax. Only students who spoke the majority language at home were included. The syntax tasks were sentence comprehension and sentence repetition with relatively short sentences with easy vocabulary, but difficult syntactic structures.
Results: Syntactic manipulations made substantial impact on accuracy levels of the syntactic tasks. This confirms that that syntactic proficiency was a central task demand. The two syntax measures were more highly correlated with each other than with decoding, vocabulary, and working memory. Vocabulary and working memory explained some variance in syntax, but the syntax tasks also explained unique variance in reading comprehension. However, the results were slightly less clear for sentence repetition than for sentence comprehension.
Conclusion: There is valid variance in grade 6 students’ proficiency with using word-order information in comprehending sentences. On the surface, sentence level tasks appear memory dependent, and they are correlated with working memory. But working memory has little to do with why sentence-level tasks are correlated with reading comprehension.