Growth in working memory and inhibition predicts literacy in English language learners: A cross-sectional and longitudinal study

Abstract

The purpose of this study was to determine whether cross-sectional and growth effects in second language (L2) literacy are related to the executive component of working memory (WM) and whether inhibition may underlie the links between WM and reading in children whose first language (L1) is Spanish. Elementary school children (grades 1, 2 and 3) were administered a battery of cognitive [WM, short-term memory (STM), random generation, rapid naming, phonological processing], vocabulary and reading measures in both Spanish (L1) and English (L2) in Year 1 and again one year later. The regression analyses showed that L2 growth in WM significantly predicted growth in L2 reading skills even when inhibition was controlled. Further, the contributions of WM to reading growth in both L1 and L2 reading were independent of cross-language skills in phonological processing, STM, oral language and naming speed. Overall, the results suggest the mental activities that underlie WM and inhibition in predictions of L2 literacy reflect independent executive processes.

Keywords:

• Phonological
• Working memory
• English language learners
• L2 reading

Special thanks are given to Cathy Lussier, Michael Orosco, Erin Bostick Mason, Valerie Perry, Joseph Rios, Elizabeth Arellano, Nicole Garcia, Alfredo Aviles, Heather Owen, Danielle Guzman-Orth, Melina Melgarejo, Elysse Bachman, Evelyn Flores, Paula Aisemberg and Steve Gómez for the data collection. Special appreciation is given to the Colton School District and San Bernardino County Superintendent of Schools and Jose Espinoza and Yolanda Cabrera for their extensive administrative support.

This is the second year of a longitudinal study funded by a grant [grant number R324A090092] from the US Department of Education, Cognition and Student Learning, Institute of Education Sciences, awarded to the author.

Special thanks are given to Cathy Lussier, Michael Orosco, Erin Bostick Mason, Valerie Perry, Joseph Rios, Elizabeth Arellano, Nicole Garcia, Alfredo Aviles, Heather Owen, Danielle Guzman-Orth, Melina Melgarejo, Elysse Bachman, Evelyn Flores, Paula Aisemberg and Steve Gómez for the data collection. Special appreciation is given to the Colton School District and San Bernardino County Superintendent of Schools and Jose Espinoza and Yolanda Cabrera for their extensive administrative support.

This is the second year of a longitudinal study funded by a grant [grant number R324A090092] from the US Department of Education, Cognition and Student Learning, Institute of Education Sciences, awarded to the author.

Notes

¹ There is some debate as to whether the backward digit span test better captures WM than STM. As suggested by Colom, Flores-Mendoza, Quiroga, and Privado (2005) and Engle et al. (1999), however, the numbers reversed task is assumed to be mainly a short-term processing capacity measure. As found by Colom et al. (2005), forward and backward span measures are similar measures of STM storage (see p. 1010, for discussion). Further, as stated by Engle et al. (1999) “Rosen and Engle (1997) showed that the backward and forward word task displayed similar effects of phonological similarity … suggesting that a simple transposition of order would be insufficient to move a task from the STM category to the WM category” (p. 314). More specifically, Rosen and Engle (1997) stated that “We found no evidence that backward recall required more complex recall than did forward recall. Also, we found that participants used a phonological code for both forward and backward serial recall” (p. 46). In addition, Swanson, Mink, and Bocian (1999) found that with young children at risk for learning problems that both forward and backward digits loaded on the same factor as phonological processing (see Table 4). Thus, because of the excellent model fit, we viewed the backward span task for this study as a measure of phonological STM.

² The disadvantage of focusing on only two waves of data is that we cannot tell the shape of each child's individual growth trajectory. However, there are advantages ascribed to latent change models that focus on two waves of data. For example, one advantage is that they accommodate a large number of variables while not requiring the meeting the assumption of convergence and factor invariance (see Hertzog, Dixon, Hultsch, & MacDonald, 2003, p. 758, for discussion). Thus, growth in our study was narrowly defined as “incremental” changes in performance.

³ There is some debate as to whether the backward digit span test better captures WM than STM. As suggested by Colom, Flores-Mendoza, Quiroga, and Privado (2005) and Engle et al. (1999), however, the numbers reversed task is assumed to be mainly a short-term processing capacity measure. As found by Colom et al. (2005), forward and backward span measures are similar measures of STM storage (see p. 1010, for discussion). Further, as stated by Engle et al. (1999) “Rosen and Engle (1997) showed that the backward and forward word task displayed similar effects of phonological similarity … suggesting that a simple transposition of order would be insufficient to move a task from the STM category to the WM category” (p. 314). More specifically, Rosen and Engle (1997) stated that “We found no evidence that backward recall required more complex recall than did forward recall. Also, we found that participants used a phonological code for both forward and backward serial recall” (p. 46). In addition, Swanson, Mink, and Bocian (1999) found that with young children at risk for learning problems that both forward and backward digits loaded on the same factor as phonological processing (see Table 4). Thus, because of the excellent model fit, we viewed the backward span task for this study as a measure of phonological STM.

⁴ WM measures administered in English and Spanish for children have been shown to load on the same latent factor (Swanson, Sáez, Gerber, & Leafstedt, 2004; Swanson, Orosco, Lussier, Gerber, & Guzman-Orth, 2011). Individual differences that emerge on each measure are related to ease of access within the preferred language, and not a language-specific cognitive system.

⁵ The disadvantage of focusing on only two waves of data is that we cannot tell the shape of each child's individual growth trajectory. However, there are advantages ascribed to latent change models that focus on two waves of data. For example, one advantage is that they accommodate a large number of variables while not requiring the meeting the assumption of convergence and factor invariance (see Hertzog, Dixon, Hultsch, & MacDonald, 2003, p. 758, for discussion). Thus, growth in our study was narrowly defined as “incremental” changes in performance.