The codevelopment of reading and attention from middle childhood to early adolescence: A multivariate latent growth curve study.

Abstract

Attention skills are strong cross-sectional predictors of reading comprehension from childhood through to adolescence. However, less is known about the developmental relations between these two domains across this period. This study examined the codevelopment of reading and attention in a community sample of 614 Australian school students (50% female). Reading and attention were assessed at ages 8, 10, 12, and 14. Results of univariate latent growth models demonstrated, on average, curvilinear trajectories for reading in which rapid growth across younger age spans decelerates as children reach adolescence. By contrast, attention skills remained relatively stable on average. Significant negative correlations were observed between the intercept and slope factors in separate reading (r = –.62) and attention models (r = –.39) suggesting compensatory growth patterns in which poorer performing students in both domains at age 8 have steeper trajectories than their higher performing peers. A comparison of a multivariate latent growth model and an autoregressive latent trajectory model with structured residuals (ALT-SR) examined the interrelatedness of development in reading and attention. Both between-individual and within-individual cross-domain parameters showed reading and attention to be positively related at Grade 3, indicating an association between better attention and higher reading achievement at age 8. However, there was little evidence for interrelated growth across domains in this sample. The results contribute to theories which explain whether and how multiple cognitive domains codevelop over a substantial period of childhood and early adolescence.

Matthew Effects in Reading and Mathematics: Examining Developmental Patterns in Population Data

“…While these results run counter to public narratives of educational decline (e.g. Kelly, 2021) they do align with prior empirical research which has found compensatory growth in reading (Larsen et al, 2022;Ricketts et al, 2020;Shin et al, 2013;Wang et al, 2017) and math (Larsen et al, 2022;Neuendorf et al, 2020;Shanley, 2016;Wang et al, 2017) across similar age spans to that examined by this study. Nonetheless, our finding of compensatory growth in NAPLAN assessments does not necessarily indicate that Matthew effects will never arise in the academic domains, and over the age span investigated by this study.…”

“…Compared with studies examining primary school grades, fewer studies have examined reading development from middle childhood through to adolescence. Across these later age spans compensatory growth patterns are more commonly identified (Larsen et al, 2022;Ricketts et al, 2020;Shin et al, 2013;Wang et al, 2017). These inconsistent results may be partly attributed to the variety of measures used to capture "reading" and the different age ranges of samples under investigation (Parrila et al, 2005): it is possible that different subcomponents of reading follow different developmental patterns and these may be obscured if assessments of several skills are aggregated, particularly in the early school grades (Paris, 2006).…”

“…Cumulative (Bast & Rietsma, 1997;Bodovski & Farkas, 2007;Little et al, 2020), compensatory (Scammacca et al, 2020;Shanley, 2016;Vogt et al, 2020;Wang et al, 2017) and stable growth patterns (Little et al, 2020) have been reported in studies of mathematics development in different subdomains, and across different age spans in primary school, mostly from kindergarten through fifth grade. Similarly, studies that examine growth from middle childhood to adolescence have identified both cumulative (Murayama et al, 2013;Muthen & Khoo, 1998;Shin et al, 2013), compensatory (Larsen et al, 2022;Neuendorf et al, 2020) and stable growth patterns (Shanley, 2016). It is possible that, similar to studies examining different subcomponent skills of reading, these varied conclusions might be driven by the age spans examined, and the specific mathematical skills that are assessed: different component skills may show different growth patterns.…”

Do student differences in reading enjoyment relate to achievement when using the random-intercept cross-lagged panel model across primary and secondary school?

“…The analyses covaried for sex (males = 0 and females = 1) and socioeconomic status (SES). SES was represented by a single factor score, as detailed in [22]. Among the five items included on the factor were the highest level of education and occupation reported by both parents and the family home's neighbourhood SES.…”

“…It is the cycle from poor skills to limited enjoyment that manifests into the Matthew Effect, wherein the weaker students fall further behind over time on account of less leisure reading, while the stronger performing students have their skills subsequently reinforced by more leisure reading, which escalates into increasingly stronger achievement over time. There are claims that the Matthew Effect perpetuates throughout schooling, but the effects are arguably short lived, as is evidenced in the literature [22] and detailed below.…”