Meta-analysis of the relation between children’s grammatical skills and mathematical learning (MetaGraM)

• Applicants: Dr. Alexander Röhm; Dr. Eva Wimmer
• Subject Area: Developmental and Educational Psychology ; General and Domain-Specific Teaching and Learning
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 554000534

In contrast to other linguistic skills, grammatical skills are attributed a more significant and longer-term impact on children’s mathematical learning. While children with mathematical learning difficulties (MD) show deficits in their grammatical skills, children with developmental language disorder (DLD) often display weak mathematical skills. An own systematic review of 95 empirical studies with children from preschool to elementary school age was conducted to shed light on this relation. The results remain descriptive due to the large heterogeneity of studies in terms of sample- and task-specific characteristics including potentially interacting third factors (e.g., age, IQ, working memory). The aim of MetaGraM is to examine the relation between children’s grammatical and mathematical skills based on the data from the previous systematic review in a meta-analysis. The reported results concerning specific grammatical and mathematical skills will be merged into total results with general levels of significance and effect sizes for the total of included studies as well as the specific subsamples of typically developing children (TD), children with DLD, and children with MD. The work programme consists of five work packages (WP). WP1 includes the study selection and follow-up screening. WP2 and WP3 realize a detailed and quality-controlled coding procedure to extract relevant statistical values and information from the studies for each relevant subsample. In WP4 the collected data will be prepared for the main analyses by determining and unifying general effect sizes and the calculation of standardized mean differences following Hedges’ g. The estimated effects will then be analyzed using random effects models (including restricted maximum-likelihood [REML] method) for all studies as well as the subsamples (TD, DLD, MD). The observed effect sizes and variances will be further investigated regarding the influence of potential moderators in meta regression models involving moderators (as IV) from sample-specific variables (e.g., IQ, working memory skills, target language, socioeconomic status) as well as task-specific variables (e.g., study design, type/specificity of grammatical or mathematical skill) for the estimated effect sizes (as DV). The main analyses will be followed-up by examination of possible publication bias. WP 5 serves to publish and disseminate the project results. The findings from MetaGraM will clarify the relation between grammatical and mathematical skills in greater detail on a valid statistical basis and enable subsequent evidence-based research and implications to foster children’s mathematical learning.

• DFG Programme: Research Grants