Preschool predictors of later reading comprehension ability: a systematic review

This Campbell systematic review examines the relationships between skills in preschool and later reading comprehension. The review summarizes evidence from 64 longitudinal studies that have observed these relationships. Code‐related skills in preschool (e.g., phoneme awareness and letter knowledge) are indirectly related to reading comprehension via word decoding. Linguistic comprehension is directly related to reading comprehension skills. Code‐related skills and linguistic comprehension were strongly related. Moreover, language comprehension was more important for reading comprehension in older readers than in younger readers. Plain language summary Preschool language skills are associated with better reading comprehension at school The evidence suggests that successful instruction for reading comprehension should target a broad set of language skills. The review in brief A variety of language skills related to both language comprehension (e.g., vocabulary and grammar) and code‐related skills (e.g., phonological awareness and letter knowledge) is important for developing decoding skills and, in turn, reading comprehension in school. Thus, reading comprehension instruction is more likely to be successful if it focuses on a broad set of language skills. What is this review about? Determining how to provide the best instruction to support children's reading comprehension requires an understanding of how reading comprehension actually develops. To promote our understanding of this process, this review summarizes evidence from observations of the development of language and reading comprehension from the preschool years into school. The main outcome in this review is reading comprehension skills. What is the aim of this review? This Campbell systematic review examines the relationships between skills in preschool and later reading comprehension. The review summarizes evidence from 64 longitudinal studies that have observed these relationships. Understanding the development of reading comprehension and its precursors can help us develop hypotheses about what effective instruction must comprise to facilitate well‐functioning reading comprehension skills. These hypotheses can be tested in randomized controlled trials. What studies are included? This review includes studies that observe the relationship between preschool language and code‐related skills and later reading comprehension. A total of 64 studies were identified, all of which were included in the analysis. However, several of them suffered from considerable attrition, used convenience sampling, included a selected sample and failed to report on important study and sample characteristics. The studies spanned 1986 to 2016 and were mostly performed in the USA, Europe and Australia. What are the main findings of this review? Code‐related skills in preschool (e.g., phoneme awareness and letter knowledge) are indirectly related to reading comprehension via word decoding. Linguistic comprehension is directly related to reading comprehension skills. Code‐related skills and linguistic comprehension were strongly related. Moreover, language comprehension was more important for reading comprehension in older readers than in younger readers. What do the findings of this review mean? These results show that a broad set of language skills is important in developing reading comprehension. The results also suggest that successful instruction for reading comprehension should target a broad set of language skills. In future studies, the effectiveness of instruction that targets such a set must be tested in randomized controlled trials. Additionally, future longitudinal studies should address issues of reliability, missing data and representativeness. How up‐to‐date is this review? The review authors searched for studies up to 2016. This Campbell systematic review was published in December 2017. Executive summary/Abstract Background Knowledge about preschool predictors of later reading comprehension is valuable for several reasons. On a general level, longitudinal studies can aid in generating understanding and causal hypotheses about language and literacy development, both of which are crucial processes in child development. A better understanding of these developmental processes may guide the establishment of effective instruction and interventions to teach reading comprehension that can later be tested in randomized controlled trials. Knowledge about preschool precursors for reading comprehension skills can also aid in developing tools to identify children at risk of reading difficulties. Objectives The primary objective for this systematic review is to summarize the available research on the correlation between reading‐related preschool predictors and later reading comprehension skills. Search methods We developed a comprehensive search strategy in collaboration with a search information retrieval specialist at the university library. The electronic search was based on seven different databases. We also manually searched the table of contents of three key journals to find additional references. Finally, we checked the studies included in two previous systematic reviews. Selection criteria The included studies had to employ a longitudinal non‐experimental/observational design. To avoid the overrepresentation of participants with special group affiliation (e.g., participants with learning disabilities or second language learner status), we chose studies that included either a sample of typically developing children or an unselected cohort. Data collection and analysis The search resulted in 3285 references. After the duplicates were removed, all remaining references were screened for inclusion and exclusion. A total of 64 studies met the eligibility criteria. The analysis was conducted in two steps. First, the predictive relation between the abilities assessed in preschool and later reading comprehension skills was analyzed using Comprehensive Meta‐analysis (CMA) software. Second, we used the correlation matrices in the included studies to further explore these relations by means of meta‐analytic structural equation modeling. Results First, analyses of bivariate correlations showed that all the included predictors, except for non‐word repetition, were moderately to strongly correlated with later reading comprehension in the bivariate analyses. Non‐word repetition had only a weak to moderate contribution to later reading comprehension ability. To explain the between‐study variation, we conducted a series of meta‐regression analyses. Age at time of reading assessment could predict variations between studies in correlations related to the code‐related predictors. Second, meta‐analytic structural equation modeling showed a significant indirect effect of code‐related skills on reading comprehension via consecutive word recognition. Third, there was a strong relationship in preschool between language comprehension and code‐related skills. Language comprehension had a moderate direct impact on reading comprehension. As hypothesized, this impact increased with age, and linguistic comprehension becomes more important for reading comprehension when children master decoding. Moreover, the overall individual variance in reading comprehension explained by the model was 59.5%; that of consecutive word recognition was 47.6%. Authors’ conclusions Overall, our findings show that the foundation for reading comprehension is established in the preschool years through the development of language comprehension and code‐related skills. Code‐related skills and decoding are most important for reading comprehension in beginning readers, but linguistic comprehension gradually takes over as children become older. Taken together, these results suggest a need for a broad focus on language in preschool‐age children.