Poster #9 - A Meta-Analysis Examining Relations between Patterning, Cognition, and Math and Reading Achievement

• In Event: Poster Session 06

Abstract

Patterning, or the ability to detect regularities in a series of elements, is central to human cognition and development (Miller et al., 2016). A growing literature examines links between patterning, cognition, and math or reading achievement, with studies often reporting positive associations between patterning and formal academic skills (Burgoyne et al., 2017). This work highlights the potential importance of early pattern knowledge for enhancing achievement outcomes. Despite this increased attention on patterning, at least two critical gaps remain in the literature: (1) Reported pattern-achievement relations are often inconsistent, and the magnitude of relations between patterning, cognition, and achievement outcomes remain unclear. For example, some correlational research indicates pattern skills relate to certain aspects of math knowledge but not others (Fyfe et al., 2017). (2) Theoretical frameworks and the mechanisms linking patterning and achievement outcomes remain unclear. In particular, it is of interest whether the relations between patterning and achievement outcomes remain even after accounting for shared variance between domain-general cognitive skills (e.g., working memory) and also whether the relations between patterning and achievement outcomes are moderated by key factors (e.g., age, pattern task).

We conducted a systematic search of the literature and identified 45 empirical papers with relevant data. Coding is underway and we plan to conduct one-stage meta-analytic structural equation modeling (MASEM) analyses based on these 45 articles to clarify relations between patterning, cognitive skills, and achievement outcomes. Our primary aim is to determine the magnitude of associations between patterning and (a) math outcomes, (b) reading outcomes, and (c) cognitive skills including executive function, intelligence, verbal skills, spatial skills, and relational reasoning. We will also examine potential moderators of these associations, including factors related to pattern tasks (i.e., pattern type or pattern activity), certain aspects of math or reading (e.g., procedural versus conceptual math skills), and demographic factors including age. Finally, we will determine the strength of (a) pattern-math and (b) pattern-reading associations after accounting for cognitive covariates.

We will conduct MASEM analyses because, unlike the traditional correlational meta-analysis that can only examine correlations between two variables, researchers can use MASEM and extracted correlation matrices from reviewed studies to create a pooled correlation matrix to simultaneously model and test interrelations between more than two variables. That is, MASEM enables researchers to evaluate more complex relations between variables, to determine hypothesized models using model fit indices including SEM, and to directly test whether a variable moderates a specific path. For these reasons, MASEM is increasingly considered a preferred meta-analytic technique for more complex model testing (Jak & Cheung, 2020).

We expect to find (1) positive, moderate pattern-math associations, with moderation analyses indicating that pattern-math relations are significant only when math outcomes measure procedural, but not conceptual knowledge; (2) positive, weak pattern-reading associations; (3) positive, moderate associations between domain-general cognitive skills; (4) after accounting for effects of cognitive covariates, pattern-math, but not pattern-reading, associations will remain. These findings will have theoretical and practical implications for elucidating the role that patterning plays in formal academic achievement.

Authors

• Giulia Alexandra Borriello, Indiana University - Bloomington
  Presenting Author

• Peng Peng, University of Texas at Austin

• Tongyao Zhang, Indiana University

• Emily Fyfe, Indiana University - Bloomington