The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysis

Kate E Mooney; Stephanie L Prady; Mary M Barker; Kate E Pickett; Amanda H Waterman

doi:10.1371/journal.pone.0260788

. 2021 Dec 2;16(12):e0260788. doi: 10.1371/journal.pone.0260788

The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysis

Kate E Mooney ^1,^*, Stephanie L Prady ¹, Mary M Barker ², Kate E Pickett ¹, Amanda H Waterman ^3,⁴

Editor: Muhammad Shahzad Aslam⁵

PMCID: PMC8639069 PMID: 34855871

Abstract

Background and objective

Working memory is an essential cognitive skill for storing and processing limited amounts of information over short time periods. Researchers disagree about the extent to which socioeconomic position affects children’s working memory, yet no study has systematically synthesised the literature regarding this topic. The current review therefore aimed to investigate the relationship between socioeconomic position and working memory in children, regarding both the magnitude and the variability of the association.

Methods

The review protocol was registered on PROSPERO and the PRISMA checklist was followed. Embase, Psycinfo and MEDLINE were comprehensively searched via Ovid from database inception until 3^rd June 2021. Studies were screened by two reviewers at all stages. Studies were eligible if they included typically developing children aged 0–18 years old, with a quantitative association reported between any indicator of socioeconomic position and children’s working memory task performance. Studies were synthesised using two data-synthesis methods: random effects meta-analyses and a Harvest plot.

Key findings

The systematic review included 64 eligible studies with 37,737 individual children (aged 2 months to 18 years). Meta-analyses of 36 of these studies indicated that socioeconomic disadvantage was associated with significantly lower scores working memory measures; a finding that held across different working memory tasks, including those that predominantly tap into storage (d = 0.45; 95% CI 0.27 to 0.62) as well as those that require processing of information (d = 0.52; 0.31 to 0.72). A Harvest plot of 28 studies ineligible for meta-analyses further confirmed these findings. Finally, meta-regression analyses revealed that the association between socioeconomic position and working memory was not moderated by task modality, risk of bias, socioeconomic indicator, mean age in years, or the type of effect size.

Conclusion

This is the first systematic review to investigate the association between socioeconomic position and working memory in children. Socioeconomic disadvantage was associated with lower working memory ability in children, and that this association was similar across different working memory tasks. Given the strong association between working memory, learning, and academic attainment, there is a clear need to share these findings with practitioners working with children, and investigate ways to support children with difficulties in working memory.

Introduction

Working memory is defined as the ability to store and process a limited amount of information over short time periods to support ongoing cognitive activities [1,2]. Working memory is also part of the broader construct of ‘executive function’; an umbrella term that encompasses the processes responsible for purposeful and goal-directed behaviour [3,4].

Working memory is essential for successful engagement in classroom activities [5], including the ability to remember and follow directions and instructions, and to engage effectively with problem-solving [6,7]. In mathematics, working memory is required to hold number combinations in mind [8] and when reading, working memory is required to keep relevant speech sounds in mind, match them up with corresponding letters, and then combine them to read words [9,10]. Indeed, working memory is positively associated with improved performance on school-based tests of English, Mathematics, and Science [11–13], and meta-analyses have found associations between working memory and mathematical performance [14,15], broad reading abilities [16], and reading comprehension ability [17]. In addition, working memory ability may underlie many broader cognitive abilities [18].

Given the importance of working memory for children’s learning and educational attainment, it is vital to understand how working memory works and what factors might influence its development. One such factor is socioeconomic position, referring to the social and economic factors that influence the positions individuals or groups hold within the structure of a society [19]. Socioeconomic position has been shown to influence multiple developmental outcomes, including educational attainment [20]. Socioeconomic gaps are also evident in children’s receptive language, general locomotor skills, and general cognitive abilities as early as 22 months old [21,22] and in school readiness, verbal ability, and spatial ability at ages 3 and 5 [23]. Ethnicity is also an important factor to consider within the context of socioeconomic position and developmental outcomes, as minority ethnic groups tend to experience both lower levels of socioeconomic position [24] and educational attainment [25]. Given the very strong associations that working memory has with broad cognitive abilities [18], and with educational attainment [11–17], it may provide a potential pathway for understanding socioeconomic inequalities in children’s outcomes. A better understanding of the association between socioeconomic position and working memory may therefore provide an understanding of one of the pathways by which socioeconomic disadvantage negatively impacts upon children’s educational attainments, and a potential route for reducing socioeconomic inequality.

Socioeconomic disadvantage is hypothesised to influence child outcomes negatively through experiences of stress and lack of access to resources [26,27]. For example, early childhood poverty is associated with increased allostatic load, a measure of physiological stress [28], and lower family income is associated with lower levels of cognitive stimulation in the home environment [29]. Whilst socioeconomic disadvantage may negatively influence child development via these suggested mechanisms, it may also be the case that enhanced social position means provides more positive enrichment and opportunities–resulting in better child development [30]. Indeed, transactional models posit that associations between genes, cognitive development, and academic outcomes are more strongly related in more advantaged socioeconomic circumstances [31,32]. Socioeconomic position could therefore influence children’s working memory via any or all of these mechanisms.

However, there is disagreement about whether or not socioeconomic disadvantage does affect working memory. Some studies have shown no link between socioeconomic position and working memory [33–35]. In contrast, other studies have found that socioeconomic disadvantage is associated with significant impairments in working memory ability [36–38]. The need to understand the precise association between socioeconomic position and working memory and the lack of consensus regarding this association indicates that a systematic review and meta-analysis of relevant studies is necessary.

No study has systematically synthesised the literature investigating the association between child socioeconomic disadvantage and working memory. Lawson, Hook and Farah [2018] investigated the association between executive functions and socioeconomic status, finding a significant, but small, association across 25 studies [39]. They found that the association was not moderated by the age of the sample or the socioeconomic indicator used. Regarding working memory specifically, Lawson et al [2018] did briefly report some follow-up analyses looking at the different components of executive function using a sub-set of twelve studies, and found a similar level of association between socioeconomic position and working memory. However, they only included studies reporting a Pearson’s r correlation. Studies that investigate socioeconomic differences in working memory often categorise children into “high” and “low” socioeconomic groups, meaning that many relevant studies would be excluded from this analysis.

Further, Lawson et al. [2018] combined all measures of working memory into one summary score for their meta-analysis. A prominent model of working memory, the Multicomponent Model, differentiates between different components of working memory, with a central executive that acts as an attentional control system, and two sub-systems that act as simple storage components [1,40]. These two storage components represent different modalities, where the phonological loop stores verbal information and the visuospatial sketchpad stores visual, spatial, and haptic information. Other influential models see working memory as a more unified construct and do not support the idea of separable, specialised working memory components [41,42]. However, many standardised measures of working memory (e.g., the Working Memory Test Battery for Children–[43]) contain tasks that differ by modality presentation (e.g., verbal vs. visuospatial stimuli) or by whether the information needs simply to be stored (e.g., forward digit span) or to be manipulated in some way (e.g, backward digit span). Further, neurocognitive studies have indicated that activities requiring attentional control are more uniquely associated with brain activation in the prefrontal cortex [44], whereas passive storage is related to activation within different networks, such as Broca’s and Wernicke’s area and the right hemisphere [45].

Given that socioeconomic disadvantage has been shown to have different patterns of association with particular aspects of cognition [33,38] and that distinct components of working memory are associated with specific underlying neurological structures, socioeconomic disadvantage may also have specific associations within distinct components of working memory. Indeed, previous studies have shown that the magnitude of the association between socioeconomic disadvantage and working memory does change dependent on whether the task material is verbal or visuospatial [34,46], and how working memory capacity is measured [47]. Further, some researchers have argued that simple storage may be more reliant on knowledge structures, which in turn are related to crystallized intelligence, and therefore may be more sensitive to the effects of socioeconomic disadvantage than attentional control, which is related more to fluid intelligence [10]. An increased understanding of the association between socioeconomic position and working memory, and how that association might vary between different aspects of working memory, is important for informing educational and clinical practitioners working with children from disadvantaged backgrounds. If poor working memory skills are part of the pathway by which socioeconomic position affects learning and academic attainment, then building in support mechanisms that specifically target this problem will be beneficial and contribute to improving outcomes for these children [8].

The current review therefore investigates the relationship between socioeconomic position and different components of working memory in children aged up to 18 years across a large number of studies using as wide a range of outcome variables as possible, and reports both the magnitude and the variability of these associations. For the purposes of this review we hereafter refer to functions within working memory relating to the processing of information requiring attentional control or executive control as ‘complex working memory’, and to functions that reflect passive storage of information as ‘simple working memory’ [48].

Methods

Protocol, registration, and reporting standards

The review protocol is registered on PROSPERO (CRD number: CRD42019134936). We used the PRISMA checklist to ensure complete and transparent reporting of methods in this review (available in supplementary online materials 4) [49].

Study eligibility criteria

Population, exposure, and outcome

We used the Population, Exposure, Outcome (PEO) framework to design the inclusion criteria [50]. The population were typically developing children aged 0–18 years old. The exposure was socioeconomic position (SEP), and we included studies with any indicators of it (e.g. parental occupation, parental education, family income, area deprivation, subjective assessment of wealth, etc.). The outcome was working memory performance, defined as any behavioural task that quantified a child’s working memory performance (e.g. Forwards Digit Recall, Backwards Digit Recall, Counting Recall, any ‘two-back’ task, Corsi). As the outcome was only working memory, and not any other cognitive or executive function tasks, studies that only reported a combined composite score on executive function were not eligible.

Study designs

Studies were eligible for inclusion if they used any observational design (cross-sectional and longitudinal), or any intervention design if they reported socioeconomic position and working memory at baseline, prior to the intervention. Studies had to provide quantitative data on the association between socioeconomic position and working memory, so qualitative studies were excluded. Although the protocol indicated that we would conduct additional searches, many studies were identified in the initial search and so only published studies were eligible for inclusion and additional searches (e.g. of grey literature) were not attempted.

Study inclusion criteria

Studies were included if they met all of the following criteria: (a) they provided data on any indicator of socioeconomic disadvantage, (b) they reported disadvantage at the individual or group level, and compared individuals or groups on that measure of disadvantage, (c) they measured performance on at least one behavioural task of working memory and reported the results quantitatively, (d) they reported data for typically developing children aged between 0–18, (e) the study was reported in the English language, (f) the study was of any observational design, or baseline characteristics if an intervention, and (g) the study was published in a peer reviewed journal.

Search and selection procedures

We searched Embase, Psycinfo and MEDLINE via Ovid to identify published articles from database inception until 3^rd June 2021. The search strategy combined key terms with a search filter that used the PROGRESS acronym to filter for equity-focused studies [51,52]. The filter is validated in Embase and MEDLINE, and was translated for use in PsycInfo. The equity filter was combined with terms and subject headings to identify ‘working memory’ abilities in ‘children’. The basic search strategy was: (search filter for equity studies) AND (subject headings) OR (“working memory”.ti,ab. OR “executive function*”.ti,ab. OR “short?term memory.ti,ab.”) AND (subject headings) OR (child* OR infant OR school child* OR adolescen* OR preschool* OR pre-school* OR boy* OR girl* OR young people OR teenager* OR teen* OR youth*.mp.). The full search strategy for Embase is provided in supplementary online materials 1.

Data extraction

The following data was extracted using a previously piloted data extraction form: location of the study, number of participants, and participant sociodemographics (gender, age range and mean, and ethnicity), exposure details (the indicator of SEP), and the measurement of the outcome (working memory). Information was extracted on ethnicity since it can be associated with socioeconomic position [24]. If information regarding participant ethnicity was not available, then data on language spoken was extracted instead if it was provided. If a study did not report information for any of the details, this was marked “NR” (not reported).

Risk of bias in individual studies

Risk of bias was assessed using one of two tools: cross-sectional studies were assessed using the AXIS appraisal tool [53], and longitudinal studies were assessed using The National Institutes of Health (NIH) Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies [54]. Risk of bias was assessed at the study level. Particular attention was paid to three key factors in both tools: (1) The selection of a defined target population with reference to the population’s socio-demographics, with a detailed sampling frame and selection process; (2) the measurement or consideration of screening to categorise children as ‘typically developing’ in the inclusion criteria for the study; (3) the measurement of working memory using a validated and or referenced task.

If a study met all three of these conditions, and successfully met the majority of the criteria from the relevant quality assessment tool, it was labelled as low risk of bias. Studies that did not meet any of the above three conditions were labelled as high risk of bias. If a study only met one or two of the three conditions, then the context of the study and other criteria within each risk of bias tool were considered before risk of bias was assigned.

Validity and reliability of review

The first reviewer (KM) screened all eligible abstracts and full texts; and a second (MB) screened a random 10% of excluded abstracts and full texts. KM extracted all data and then MB checked the data extraction and risk of bias assessments for 50% of all included studies. Agreement between the reviewers was considered to be acceptable if it was at least 90% at all stages, and any disagreements were resolved through discussion.

Data synthesis

We used two methods of numerical data synthesis: random-effects meta-analyses to combine studies with eligible effect estimates, and a Harvest plot to synthesise findings from otherwise eligible studies without effect size estimates or that used a composite measure of working memory. Fig 1 provides a summary of how studies were selected into each data synthesis method.

Two meta-analyses were conducted by the type of working memory: (1) simple working memory and (2) complex working memory. Studies were therefore included in a meta-analysis if they reported a useable (or convertible) unadjusted effect size between socioeconomic position and working memory on ≥1 task(s) of working memory that could be conceptualised as either simple working memory, or complex working memory. We also conducted subgroup estimation within both meta-analyses, depending on whether the task modality was verbal or visuospatial. A small number of studies combined verbal and visuospatial task modalities and in order to include as many studies as possible within this analysis, we still included those studies that had a combined score, as long as they had separate measurements of simple working memory and complex working memory. Cohen’s d effect sizes were calculated for all studies that provided mean scores across two groups of SEP. Not all studies provided mean scores and we converted correlations to Cohen’s d effect size where possible, using formulae provided by Borenstein et al., (2009) [55]. This therefore means that the meta-analysis represents a metric comparing lower socioeconomic position to higher socioeconomic position groups.

The Harvest plot contained studies that were deemed ineligible for the meta-analysis. A small number (n = 8) of studies used composite measures of working memory (with both simple and complex working memory tasks). We decided to include these for the sake of completeness, even though one of our key aims was to investigate the association between socioeconomic position and different components of working memory. The Harvest plot therefore included studies that reported what we categorised as a composite working memory score: (i) either a combined score on tasks of both ≥1 simple working memory task(s) and ≥1 complex working memory task(s) or (ii) a single task of working memory including elements of both simple and complex working memory. Further, we also included studies which reported (iii) an effect size that was adjusted for other factors and so could not be appropriately used in meta-analysis (e.g. regression analysis).

Data synthesis: Meta-analytic methods

Investigation of heterogeneity

Heterogeneity was calculated for each meta-analysis using the I² statistic, and 95% prediction intervals. The I² statistic is a statistical test where 0% to 40% might represent unimportant heterogeneity, 30% to 60% moderate heterogeneity, 50% to 90% substantial heterogeneity, and 75% to 100% considerable heterogeneity [56]. Prediction intervals present the heterogeneity in the same metric as the original effect size measure and illustrate the range of true effects that can be expected in future settings [57]. To investigate sources of heterogeneity, we also undertook sensitivity analyses and meta-regression analyses.

Sensitivity analyses

The majority of studies reported two or more effect sizes that were eligible for the meta-analyses (70%), e.g. SEP-working memory correlations for the same individuals at different ages or time points [58], and these estimates are statistically dependent. We first averaged the effect sizes to give one effect size per study, however, this may result in a loss of potentially important information, improper sampling variance, or a higher probability of type-2 errors [59–61]. As a sensitivity analysis we re-estimated the meta-analyses using the Robust Variance Estimation (RVE) method, which accounts for statistically dependent effect sizes [62], and compared the results to the main analyses where effect sizes were averaged. As RVE only provides an overall summary effect size, neither subgroup analysis by verbal and visuospatial or estimation of heterogeneity parameters were possible for this sensitivity analysis. We also conducted a sensitivity analysis to examine the potential effect the inclusion or exclusion of a single study with a very large effect size.

Meta regression analyses

Meta-regression allows the effect of both continuous and categorical characteristics on the estimated effect size to be investigated. It estimates whether the association of interest (socioeconomic position and working memory) is associated with an investigated characteristic, where a significant p value indicates evidence that it is a significant moderator [63].

We tested moderation of the association between socioeconomic position and working memory with three characteristics as pre-specified on PROSPERO: (a) the type of socioeconomic indicator (whether it was a composite or single indicator), (b) the risk of bias (low or high), and (c) the task modality (verbal or visuospatial). We also tested three further post-hoc moderators; (d) the type of effect size (Cohen’s d or converted from Pearson’s r), (e) whether the effect size had been averaged from >1 estimate(s) or not, and (f) the mean age of the sample. We wanted to ensure the effect sizes were not affected by the way they were converted or combined. We tested moderation by age as enough data was available to do so, and to investigate if this could explain the heterogeneity found. If the statistical significance test was p < .05, the tested variable was considered to be a significant moderator of the association between socioeconomic position and working memory.

Publication bias

Publication bias was investigated for each meta-analysis using a funnel plot, with effect estimates plotted against standard errors of effect estimates, and Egger’s test, which estimated whether the association between study size and effect estimates was greater than expected by chance [64]. As publication bias is typically assessed per meta-analysis, this was not done by subgroup analysis (verbal vs visuospatial).

Software

We used Stata-16 and the Meta command for the computation of effect sizes, calculation of heterogeneity statistics, calculation of pooled effect sizes, meta-regressions, and producing the forest-plots [65]. For the sensitivity analysis regarding robust variance estimation, we used the Stata-16 user written command robumeta [66].

Data synthesis: Harvest plot

Studies were grouped on the Harvest plot based on whether reported findings illustrated a positive association (lower socioeconomic position and lower working memory), negative association (lower socioeconomic position and higher working memory), or no association. Outcome measures, study designs, and study quality were summarised [67]. The Harvest plot enables the reader to judge where the majority of studies lie in relation to competing hypotheses, and where the highest quality studies are. As some studies included numerous socioeconomic indicators, those that report multiple effect sizes are represented in the plot more than once. The columns represent effect sizes across composite working memory, simple working memory, and complex working memory.

Results

Study selection

The study selection process is reported according to the PRISMA STATEMENT (http://www.prisma-statement.org/) diagram.

Fig 2 shows the selection process for all included studies, a total of 64 studies were eligible for the review. The majority of these were meta-analysed (n = 36), and the remaining studies were included in the Harvest plot (n = 28).

Study characteristics

A summary of included study characteristics and risk of bias results are provided in online supplementary materials 2, and a full reference list of studies is provided in online supplementary materials 3. The 64 studies were conducted across 17 different countries, with the most frequently studied populations being the USA (34%) and Brazil (11%). The age range included children from aged 1 month to 17 years, and the majority of studies included only children under 10 years old (56%), whilst the remaining studies included a mix of ages both below and above 10 years old (44%). A variety of measures of socioeconomic position were used, with most studies including traditional measures of parental occupation, parental education and family income (80%). Other studies used measures including school socioeconomic coefficients, country specific socioeconomic indicators, and neighbourhood measures (15%), and a few studies used subjective measures of social status (5%). Many of the studies included ≥2 ethnic groups, with both ethnic majority and minority children (45%). The remaining studies either included only ethnic majority children (20%), only ethnic minority children (3%), or either did not report ethnicity at all or only reported the language spoken by the sample (30%). The majority of studies (80%) were rated as low risk of bias; studies at high risk of bias (20%) were rated as high risk either because they did not specify the socio-demographics of their population, did not have clear inclusion criteria, or used a measure of working memory that was not referenced or validated.

Meta-analyses

Summary of effects

There were 25,249 individual participants from 35 individual studies included across both meta-analyses. Results are presented firstly for the meta-analysis of simple working memory and then the meta-analysis of complex working memory. Within each meta-analysis, the subgroup analysis is presented by modality (verbal vs visuospatial).

Simple working memory. Fig 3 shows the meta-analysis of simple working memory, which included 27 studies with 14,328 participants (including 7006 from one study). The effect size and 95% CI was 0.45 (0.27 to 0.63). In the task modality subgroup analysis, the verbal estimate and its 95% CI was 0.47 (0.15 to 0.79), the visuospatial estimate 0.40 (0.23 to 0.57), and the combination of verbal and visuospatial estimate was 0.55 (0.16 to 0.94).

Complex working memory. Fig 4 shows the complex working memory meta-analysis, which included 23 studies with 20,651 participants (including 14,000 from one study). The effect size and 95% CI was 0.52 (0.31 to 0.72). In the subgroup analysis of task modality, the verbal estimate was 0.54 (0.25 to 0.83), the visuospatial estimate 0.41 (0.13 to 0.69), and the combination of verbal and visuospatial estimate 0.62 (0.42 to 0.82).

Heterogeneity

Heterogeneity was high overall. I² was 85% overall in simple working memory, with substantially higher heterogeneity in simple verbal working memory (89%) than simple visuospatial working memory (48%). I² was 87% overall in complex working memory, with again substantially higher heterogeneity in complex verbal working memory (91%) than complex visuospatial working memory (47%) (likely due to the subgroup analysis including only 4 studies). Prediction intervals were wide and overlapped with the null, indicating some uncertainty about the direction and magnitude of any effect to be expected in a new study. The 95% prediction intervals were -0.399 to 1.297 for simple working memory, and -0.407 to 1.438 for complex working memory.

Publication bias

We assessed publication bias for each of the two meta-analyses. The funnel plots in Figs 5 and 6 were both judged to be symmetrical and did not show an association between study size and study effect estimates. The Egger’s tests were both non-significant (simple working memory p = .44, complex working memory p = .93), again indicating low risk of publication bias.

Sensitivity analysis

The RVE analysis of simple working memory included 27 studies with 50 individual effect sizes. The simple working memory effect size and 95% CI was 0.44 (0.24 to 0.64). The RVE analysis of complex working memory included 23 studies with 39 individual effect sizes. The complex working memory effect size and 95% CI was 0.53 (0.30 to 0.75). As these estimates are extremely similar to when effect sizes were averaged within studies, here we have presented only the forest plots for the averaged effect sizes and the discussion focuses on the results when the effect sizes were averaged.

Removing one study [68] with substantially larger effect sizes than others in both meta-analyses (d = 2.17 in simple working memory, and d = 2.22 in complex working memory) reduced the effect sizes by approximately 0.1 (simple working memory from 0.45 to 0.37 and complex working memory 0.52 to 0.42). Removing this study also substantially reduced the heterogeneity as measured by I², from 87% to 48% in simple working memory, and from 88% to 43% in complex working memory. As the overall effect sizes were still within the bounds interpreted as “medium”, we retained the Aran-Filippetti (2013) study in all meta-analyses.

Meta regression analyses

Results from the meta-regression analysis are presented in Table 1. We conducted pre-specified moderation analyses by the task modality, and type of socioeconomic indicator; however, neither of these variables significantly moderated the association between socioeconomic position and working memory. As a post-hoc analysis, we found that age in years did not significantly moderate the association (for those studies that reported mean age in years), nor did the risk of bias of the study. We also found that whether the effect size was averaged or not did not significantly moderate the association, nor did whether the effect size was converted from Pearson’s r. However, the test was borderline significant (p = .05) for the second moderation test in simple working memory.

Table 1. Meta-regression analyses results.

	Simple working memory		Complex working memory
Regression factor	B (95% CI)	p	B (95% CI)	p
Pre-specified
Task modality (0 = verbal, 1 = visuospatial)*	-0.07 (-0.50 to 0.35)	.47	-0.26 (-0.91 to 0.38)	.42
Socioeconomic indicator (0 = single, 1 = composite)	-.11 (-0.48 to 0.27)	.58	-.00 (-0.44 to 0.43)	.97
Post-hoc
Risk of bias (0 = low risk, 1 = high risk)	-0.20 (-.60 to .21)	.36	-0.20 (-0.77 to 0.36)	.49
Effect size (0 = Cohen’s d, 1 = Converted from Pearson’s r)	-0.35 (-0.71 to -0.00)	.05	-0.18 (-0.63 to 0.26)	.41
Effect size (0 = single, 1 = averaged)	-.17 (-0.55 to 0.02)	.40	0.19 (-0.27 to 0.65)	.42
Age in years**	-.05 (-0.11 to .00)	.09	-.02 (-.07 to .03)	.43

Open in a new tab

*Three studies used combined estimates of verbal and visuospatial task modalities, and were excluded from this analysis.

*Nine studies did not report a mean age of their sample, and were excluded from this analysis.

Harvest plot

There were 28 studies included in the Harvest plot using 51 effect sizes from 12,488 individual participants. The majority of studies contributed ≥2 effect sizes (58%).

In Fig 7, The Harvest plot shows the distribution of statistically significant associations and non-statistically significant associations across composite working memory, simple working memory, and complex working memory by socioeconomic position measure. Studies only showed a positive association (increased socioeconomic position and increased working memory), or no association, so there is no column representing negative association. The abundance of studies in the composite working memory columns relative to the simple and complex working memory columns reflects that studies with composite working memory measures were not included in the meta-analyses.

Nineteen individual studies including 7826 participants provided 43 effect sizes on composite working memory. The majority of studies found composite working memory to be significantly positively associated with composite socioeconomic indicators, household wealth and parental education, and most of these studies were rated as low risk of bias. Two studies rated as low risk of bias found no association between composite socioeconomic position and working memory, and two studies rated as low risk of bias found single parent status to not be associated with verbal working memory. Eight individual studies including 7826 participants provided 14 effect sizes for simple working memory. Simple working memory was found to be associated with composite socioeconomic position indicators, household wealth, and parental education, and most of these studies were rated as low risk of bias. Only three studies found no association between socioeconomic position and simple working memory. Three individual studies including 641 participants provided four effect sizes for complex working memory. Complex working memory was found to be associated with composite socioeconomic position and household wealth in two different studies, one of which was rated as low risk of bias. The third study of complex working memory, rated as low risk of bias, found no association with composite socioeconomic position.

Overall, the Harvest plot indicates an association between socioeconomic position and different types of working memory across different indicators of socioeconomic position, that appear unrelated to risk of bias. Although there were some studies that found evidence against these hypotheses, the weight of evidence rated as low risk of bias was much more in favour of supporting the evidence for an association.

Discussion

This is the first systematic review of the association between socioeconomic position and children’s working memory abilities, with a very large sample of individual participants (n = 37,737) across two data synthesis methods. In a meta-analysis of 27 studies with 14,328 participants, socioeconomic position was associated with overall simple working memory ability with a medium effect size. In a meta-analysis of 23 studies with 20,651 participants, socioeconomic position was associated with overall complex working memory, also with a medium effect size. Furthermore, socioeconomic position was significantly associated with both verbal and visuospatial tasks within both simple and complex working memory. We also synthesized 28 studies including 12,488 participants with more diverse measures of effect using a Harvest plot, finding that most predictors of socioeconomic position are associated with working memory. The findings are consistent with literature that views socioeconomic disadvantage to be associated with impairments in working memory [36,37], and therefore does not support the view that working memory is unrelated to socioeconomic disadvantage [10,33].

We found that the magnitude of the association was similar across both the simple and complex working memory meta-analyses (d = 0.45 and d = 0.52 for simple and complex working memory, respectively). This indicates that child socioeconomic disadvantage is associated with not only difficulties in the simple storage of information, but also with the ability to process and manipulate information. This does not support the argument that simple working memory may be more sensitive to the effects of socioeconomic disadvantage than complex working memory due to being more reliant on knowledge structures [10].

We also investigated whether the magnitude of the association differed by modality (verbal and visuospatial), finding a similar magnitude of associations within the simple working memory meta-analysis (d = 0.47 and d = 0.40 for verbal and visuospatial, respectively) and the complex working memory meta-analysis (d = .54 and d = 0.41 for verbal and visuospatial working memory, respectively). We also tested this formally through meta-regression, and found that task modality did not moderate the association between socioeconomic position and working memory. Still, visuospatial working memory tended to have smaller effect sizes, and this is likely because our subgroup analyses of the visuospatial studies contained fewer effect sizes–perhaps due to difficulties with assessing visuospatial working memory in children. Future studies could examine the consistency of the association between socioeconomic disadvantage and visuospatial working memory to explore this further.

Modular working memory theories, such as the multicomponent model [1,69], propose separate components for different functions within working memory. In contrast, unitary approaches such as the attentional control model [70] and Cowan’s embedded processes model [41,71] do not support dissociable components. Our results showed a similar level of association between socioeconomic position and different components of working memory. This could be seen as evidence to support a more unitary approach to working memory. However, it may also be that the separate components of working memory are affected by socioeconomic position to a similar extent.

The findings indicated significant heterogeneity across the studies, with prediction intervals crossing the null line. However, the prediction intervals included a high upper boundary and the average effect size in both studies is medium, indicating that a significant average effect is likely to exist in future settings [57]. High heterogeneity can be due to clinical or methodological diversity, and in most cases, it is likely due to both [72]. It was difficult to ascertain the source of heterogeneity in this review as we synthesised a large number of studies, varying in both methodological and participant characteristics. The finding of high heterogeneity can be interpreted as an indication that the association between socioeconomic position and working memory is highly likely to vary across different settings and participants. Further, the prediction intervals overlapped with the null, indicating some uncertainty about the direction and magnitude, and therefore uncertainty regarding the generalizability of the effect to future studies.

We investigated some sources of the high heterogeneity through exploration of potential moderating characteristics using meta-regression [73]. We found that risk of bias did not moderate the association, where studies at high risk of bias had similar associations as those with low risk of bias. This may be because only a small proportion of meta-analyzed studies were assessed to be high risk of bias (20%). We found that child age did not moderate the association which suggests that socioeconomic disadvantage is detrimental to children’s working memory regardless of child age, and does not accumulate throughout childhood. Still, as the majority of studies in this review were cross-sectional in design, this finding warrants further validation with longitudinal studies. Finally, whether or not the socioeconomic indicator was a single item or a composite did not moderate the association. We were only able to compare the difference between single and composite indicators of socioeconomic position, as there were not enough data to explore differences between different individual indicators. This finding therefore warrants further exploration across single indicators of socioeconomic position and working memory, as this may give more insight into any causal mechanisms between disadvantage and working memory.

We also explored the influence of our data synthesis methods on the effect sizes through meta-regression. We found that the association was not moderated by whether the effect size had been averaged or not (and this was further confirmed with the RVE sensitivity analysis). Finally, there was some evidence to suggest the association between socioeconomic position and simple working memory was moderated by whether effect sizes had been converted from Pearson’s r, with smaller effect sizes for those that had been converted (B = -0.35, p = .05), although this finding did not hold for complex working memory. Studies that had been converted from Pearson’s r showing some evidence of weaker associations than those that used mean scores across two groups is relatively unsurprising since studies representing a continuum of socioeconomic position would have weaker associations than those comparing two extreme groups of socioeconomic position. This finding may therefore suggest the true association between two extreme groups of socioeconomic position and working memory is even larger than we have estimated here.

The finding that working memory is associated with socioeconomic position has important implications for educational and clinical professionals who work with children from disadvantaged backgrounds. Poor working memory ability is strongly linked to worse learning and educational outcomes. Therefore, practitioners should consider whether these children could benefit from an environment that actively scaffolds and supports children’s working memory [8].

Directions for future research

We did not systematically investigate causal or contextual factors that may mediate the association between socioeconomic position and working memory as this was not the focus of our review. Further investigation using longitudinal studies would enable exploration of the complex interplay between different factors and the effect on working memory, and we recommend some factors for future research here.

One potential moderating characteristic is ethnicity. It was not possible for us to explore ethnicity as a moderator as nearly half of the studies in this review included two or more ethnic groups, with both ethnic majority and minority children. Minority ethnic groups tend to experience higher levels of socioeconomic disadvantage [24], and it has previously been found that socioeconomic disadvantage is associated with worse working memory in ethnic minority children, whilst ethnic majority children at different levels of socioeconomic risks have similar working memory ability [74]. This therefore indicates that a difference across ethnic groups in how socioeconomic position may influence working memory may exist, and ethnicity may be a potential moderator of this association. The disadvantage faced by ethnic minority groups may exacerbate the negative association between socioeconomic position and working memory, something that could be explored more fully in future research.

As mentioned in the introduction, two key potential mediating causal factors between socioeconomic disadvantage and child development are the home learning environment and chronic stress [27–29]. Socioeconomic disadvantage may impair parents’ ability to provide home enrichment resources and activities (use of toys, books, and learning experiences), which has been found to be associated with children’s working memory [36]. Additionally, allostatic load, a biological marker of cumulative chronic stress, has been found to mediate the associations between childhood poverty and adult working memory ability [75], and this is consistent with a systematic review that found an association between early life stress and working memory [76].

A contextual factor that may induce differences is ‘stereotype threat’. Stereotype threat occurs when people are, or feel themselves to be, at risk of conforming to stereotypes about their own social group, and has been discussed as a contributing factor to the achievement gap between children of low and high socioeconomic status [77] and children from different ethnic groups [78]. Schmader, Johns and Forbes (2008) theorise that for those at risk of being negatively stereotyped about their abilities, stereotype threat increases physiological stress at the time of testing, active monitoring of performance, and efforts to suppress negative thoughts. These physiological and psychological mechanisms consume executive resources needed to perform well on cognitive tasks, including tasks of working memory [79]. Whilst the majority of studies investigating stereotype threat explicitly prime stereotypes prior to test tasking [77,80], children may still be aware of their disadvantage in a test setting without explicit priming. As socioeconomically disadvantaged children become aware of their relative disadvantage early in life [81], it seems plausible that stereotype threat may underpin some socioeconomic differences in working memory.

Finally, whether the association between socioeconomic position and working memory has a particular impact on specific areas of educational attainment is of interest, and longitudinal analyses examining working memory as a mediator between socioeconomic position and educational attainment could reveal more about these associations. Indeed, one study has found that executive function partially mediates the association between socioeconomic position and math achievement [37].

Strengths and limitations

This systematic review included a broad range of studies using a variety of methods to assess the association between socioeconomic position and working memory. The use of a comprehensive search strategy utilising the equity filter based on PROGRESS [51] allowed us to identify a large number of studies (>7000 at the initial stage). Unlike previous reviews on this topic, inclusion was not constrained to any particular estimation method, but included all studies with any quantitative measure of association between socioeconomic position and working memory. The use of the Harvest plot allowed us to include studies using any estimation method and reduces the likelihood of bias in the findings. This systematic review is the first to analyse the association between socioeconomic position working memory, and explores the association by the different components of working memory. The separation of the results into the different components of working memory allows the results to be applicable to both modular and unitary working memory models, as the summary effect sizes for each component can be considered to reflect those different components of working memory, or they can be combined to consider working memory as one construct.

As the majority of studies used cross-sectional designs, we were not able to establish causality from the associations reported in this review. However, we have highlighted potential causal factors for future studies to investigate. We converted effect size measures to a common metric, and thus the conversion into Cohen’s d therefore means that our meta-analyses analysed socioeconomic position as a dichotomous variable with two groups of socioeconomic position–which is not how socioeconomic position is actually distributed. However, the alternative would have been to exclude the studies that happened to use an alternate metric—potentially resulting in a biased sample of studies [73].

Conclusion

To conclude, this is the first systematic review specifically to investigate the association between socioeconomic disadvantage and working memory, and to analyse that relationship across different components of working memory. The results showed that socioeconomic disadvantage was associated with lower working memory ability in children, and that this association was similar across different working memory tasks. This review adds to a large body of evidence demonstrating the unjust developmental inequalities faced by children from socioeconomically disadvantaged families face [20–23]. Given the strong association between working memory, learning, and academic attainment, there is a definite need to investigate whether the pathway between socioeconomic position and working memory may explain some of the stark socioeconomic inequalities in children’s educational attainments. In addition, there is a need to share these findings with practitioners working with children, and to continue to investigate ways to support children with difficulties in working memory.

Supporting information

S1 File. Search strategy for Embase.

(DOCX)

Click here for additional data file.^{(25.3KB, docx)}

S2 File. Table of study characteristics.

(DOCX)

Click here for additional data file.^{(41.9KB, docx)}

S3 File. References for studies included in systematic review.

(DOCX)

Click here for additional data file.^{(28.6KB, docx)}

S4 File. PRISMA 2009 checklist.

(DOC)

Click here for additional data file.^{(63.5KB, doc)}

Data Availability

All relevant data are within the manuscript and its supporting information files.

Funding Statement

KEP, AHW, and SLP's involvement was supported by the National Institute for Health Research Yorkshire and Humber ARC (reference: NIHR200166). The views expressed in this publication are those of the authors and not necessarily those of the National Institute for Health Research or the Department of Health and Social Care. The work of the lead author (KEM) was supported by an ESRC White Rose Doctoral Training Partnership Pathway Award. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Baddeley A. Working memory. Curr Biol. 2010. [DOI] [PubMed] [Google Scholar]
2.Cowan N. The many faces of working memory and short-term storage. Psychon Bull Rev. 2017;24(4):1158–70. doi: 10.3758/s13423-016-1191-6 [DOI] [PubMed] [Google Scholar]
3.Anderson P. Assessment and Development of Executive Function (EF) During Childhood. Child Neuropsychol [Internet]. 2002;8(2):71–82. Available from: http://www.tandfonline.com/doi/abs/10.1076/chin.8.2.71.8724 [DOI] [PubMed] [Google Scholar]
4.Diamond A, Lee K. Interventions shown to aid executive function development in children. Science (80-). 2011;333(October):959–64. doi: 10.1126/science.1204529 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Gathercole SE, Durling E, Evans M, Jeffcock S, Stone S. Working memory abilities and children’s performance in laboratory analogues of classroom activities. Appl Cogn Psychol [Internet]. 2008. Dec 1 [cited 2021 Jan 4];22(8):1019–37. Available from: http://doi.wiley.com/10.1002/acp.1407. [Google Scholar]
6.Waterman AH, Atkinson AL, Aslam SS, Holmes J, Jaroslawska A, Allen RJ. Do actions speak louder than words? Examining children’s ability to follow instructions. Mem Cogn [Internet]. 2017. Aug 1 [cited 2020 Dec 2];45(6):877–90. Available from: https://link.springer.com/article/10.3758/s13421-017-0702-7. doi: 10.3758/s13421-017-0702-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Allen RJ, Waterman AH. How does enactment affect the ability to follow instructions in working memory? Mem Cogn [Internet]. 2015. Apr 1 [cited 2020 Dec 2];43(3):555–61. Available from: https://link.springer.com/article/10.3758/s13421-014-0481-3 [DOI] [PubMed] [Google Scholar]
8.Gathercole S, Alloway T. Working memory and learning: A practical guide for teachers. Sage; 2008. [Google Scholar]
9.Cowan N. Working Memory Underpins Cognitive Development, Learning, and Education. 2014;197–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Alloway TP, Copello E. Working memory: The what, the why, and the how. Aust Educ Dev Psychol. 2013;30(2):105–18. [Google Scholar]
11.Gathercole SE, Pickering SJ, Knight C, Stegmann Z. Working memory skills and educational attainment: Evidence from national curriculum assessments at 7 and 14 years of age. Appl Cogn Psychol. 2004;18(1):1–16. [Google Scholar]
12.Holmes J, Adams JW. Educational Psychology Working Memory and Children’s Mathematical Skills: Implications for mathematical development and mathematics curricula. 2007. [cited 2020 Nov 11]; Available from: https://www.tandfonline.com/action/journalInformation?journalCode=cedp20. [Google Scholar]
13.St Clair-Thompson HL, Gathercole SE. Executive functions and achievements in school: Shifting, updating, inhibition, and working memory. Q J Exp Psychol. 2006;59(4):745–59. [DOI] [PubMed] [Google Scholar]
14.Allen K, Higgins S, Adams J. The Relationship between Visuospatial Working Memory and Mathematical Performance in School-Aged Children: a Systematic Review. Educ Psychol Rev. 2019. doi: 10.1007/s10648-018-9458-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Peng P, Namkung J, Barnes M, Sun C. A meta-analysis of mathematics and working memory: Moderating effects of working memory domain, type of mathematics skill, and sample characteristics. J Educ Psychol [Internet]. 2016. May 1 [cited 2020 Nov 11];108(4):455–73. Available from: /record/2015-41049-001. [Google Scholar]
16.Peng P, Barnes M, Wang C, Swanson HL. A Meta-Analysis on the Relation Between Reading and Working Memory. 2017. [cited 2020 Nov 11]; Available from: 10.1037/bul0000124.supp. [DOI] [PubMed] [Google Scholar]
17.Carretti B, Borella E, Cornoldi C, De Beni R. Role of working memory in explaining the performance of individuals with specific reading comprehension difficulties: A meta-analysis. Learn Individ Differ [Internet]. 2009;19(2):246–51. Available from: 10.1016/j.lindif.2008.10.002. [DOI] [Google Scholar]
18.Gruszka A, Nęcka E. Limitations of working memory capacity: The cognitive and social consequences. Eur Manag J. 2017. Dec 1;35(6):776–84. [Google Scholar]
19.Krieger N, Williams DR, Moss NE. Measuring Social Class in US Public Health Research: Concepts, Methodologies, and Guidelines. Annu Rev Public Health [Internet]. 1997;18(1):341–78. Available from: http://www.annualreviews.org/doi/10.1146/annurev.publhealth.18.1.341. [DOI] [PubMed] [Google Scholar]
20.Rees G, Gromada A, Cuesta J, Bruckauf Z. An Unfair Start. Inequality in Children’s Education in Rich Countries. 2018. [Google Scholar]
21.Halle T, Forry N, Hair E, Perper K, Wandner L, Wessel J, et al. Disparities in Early Learning and Development: Lessons from the Early Childhood Longitudinal Study-Birth Cohort (ECLS-B) [Internet]. 2009. [cited 2020 Dec 2]. Available from: http://www.nccp.org/publications/pdf/text_851.pdf. [Google Scholar]
22.Feinstein L (London S of E. Inequality in the Early Cognitive Development of British Children in the 1970 Cohort. 2003;73–97. [Google Scholar]
23.Jerrim J, Vignoles A. Social mobility, regression to the mean and the cognitive development of high ability children from disadvantaged homes. J R Stat Soc Ser A Stat Soc. 2013;176(4):887–906. [Google Scholar]
24.Chattoo S, Atkin K. Understanding ‘race’ and ethnicity 2e: Theory, history, policy, practice. 2nd ed. Policy Press; 2019. doi: 10.1242/jcs.224147 [DOI] [Google Scholar]
25.Ethnicity Parsons C., gender, deprivation and low educational attainment in England: Political arithmetic, ideological stances and the deficient society. Educ Citizsh Soc Justice [Internet]. 2016. Jul 8 [cited 2020 Nov 25];11(2):160–83. Available from: http://journals.sagepub.com/doi/10.1177/1746197916648282. [Google Scholar]
26.Bornstein MH. Toward a model of culture↔parent↔child transactions. In: The transactional model of development: How children and contexts shape each other [Internet]. American Psychological Association; 2009. [cited 2020 Nov 25]. p. 139–61. Available from: /record/2009-05801-008. [Google Scholar]
27.Duncan GJ, Magnuson K, Votruba-Drzal E. Moving beyond Correlations in Assessing the Consequences of Poverty. Annu Rev Psychol. 2017;68:413–34. doi: 10.1146/annurev-psych-010416-044224 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Lupien SJ, King S, Meaney MJ, McEwen BS. Can poverty get under your skin? Basal cortisol levels and cognitive function in children from low and high socioeconomic status. Dev Psychopathol [Internet]. 2001. Jun [cited 2020 Dec 4];13(3):653–76. Available from: https://www.cambridge.org/core/journals/development-and-psychopathology/article/abs/can-poverty-get-under-your-skin-basal-cortisol-levels-and-cognitive-function-in-children-from-low-and-high-socioeconomic-status/44990AC6E1E448F8414FD171FE25CA5D. doi: 10.1017/s0954579401003133 [DOI] [PubMed] [Google Scholar]
29.Votruba-Drzal E. Economic disparities in middle childhood development: Does income matter? Dev Psychol. 2006. Nov;42(6):1154–67. doi: 10.1037/0012-1649.42.6.1154 [DOI] [PubMed] [Google Scholar]
30.Amso D, Lynn A. Distinctive Mechanisms of Adversity and Socioeconomic Inequality in Child Development: A Review and Recommendations for Evidence-Based Policy. Policy Insights from Behav Brain Sci. 2017;4(2):139–46. doi: 10.1177/2372732217721933 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Tucker-Drob EM, Briley DA, Harden KP. Genetic and Environmental Influences on Cognition Across Development and Context. Curr Dir Psychol Sci [Internet]. 2013. Oct 25 [cited 2021 Apr 7];22(5):349–55. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24799770 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Peng P, Kievit RA. The Development of Academic Achievement and Cognitive Abilities: A Bidirectional Perspective. Child Dev Perspect [Internet]. 2020. Mar 27 [cited 2021 Apr 7];14(1):15–20. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1111/cdep.12352. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Engel PMJ, Santos FH, Gathercole SE. Are working memory measures free of socio-economic influence? Anim Genet. 2008;39(5):561–3. doi: 10.1111/j.1365-2052.2008.01757.x [DOI] [PubMed] [Google Scholar]
34.Vandenbroucke L, Verschueren K, Ceulemans E, De Smedt B, De Roover K, Baeyens D. Family demographic profiles and their relationship with the quality of executive functioning subcomponents in kindergarten. Br J Dev Psychol. 2016. Jun 1;34(2):226–44. doi: 10.1111/bjdp.12127 [DOI] [PubMed] [Google Scholar]
35.Wiebe SA, Espy KA, Charak D. Using Confirmatory Factor Analysis to Understand Executive Control in Preschool Children: I. Latent Structure. Dev Psychol. 2008;44(2):575–87. doi: 10.1037/0012-1649.44.2.575 [DOI] [PubMed] [Google Scholar]
36.Hackman DA, Betancourt LM, Gallop R, Romer D, Brodsky NL, Hurt H, et al. Mapping the trajectory of socioeconomic disparity in working memory: Parental and neighborhood factors. Child Dev. 2014;85(4):1433–45. doi: 10.1111/cdev.12242 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Lawson GM, Farah MJ. Executive function as a mediator between SES and academic achievement throughout childhood. Int J Behav Dev. 2017;41(1):94–104. doi: 10.1177/0165025415603489 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Farah MJ, Shera DM, Savage JH, Betancourt L, Giannetta JM, Brodsky NL, et al. Childhood poverty: Specific associations with neurocognitive development. Brain Res. 2006;1110(1):166–74. doi: 10.1016/j.brainres.2006.06.072 [DOI] [PubMed] [Google Scholar]
39.Lawson GM, Hook CJ, Farah MJ. A meta-analysis of the relationship between socioeconomic status and executive function performance among children. Dev Sci. 2018;21(2):1–22. doi: 10.1111/desc.12529 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Baddeley AD. Working memory: Theories, models, and controversies. Explor Work Mem Sel Work Alan Baddeley. 2012;333–69. [DOI] [PubMed] [Google Scholar]
41.Cowan N, Morey CC, Naveh-Benjamin M. An Embedded-Processes Approach to Working Memory. In: Logie R, Camos V, & Cowan N (Eds), Working Memory: The State of the Science. Oxford University Press.; 2021. [Google Scholar]
42.Oberauer K. Towards a theory of working memory: From metaphors to mechanisms. In: Logie R, Camos V, & Cowan N (Eds), Working Memory: The State of the Science. Oxford University Press; 2021. [Google Scholar]
43.Pickering S, Gathercole SE. Working memory test battery for children (WMTB-C). Psychological Corporation; 2001. [Google Scholar]
44.D’Esposito M, Postle BR. The cognitive neuroscience of working memory. Annu Rev Psychol [Internet]. 2015. Jan 1 [cited 2021 Jan 8];66:115–42. Available from: www.annualreviews.org. doi: 10.1146/annurev-psych-010814-015031 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Chai WJ, Abd Hamid AI, Abdullah JM. Working memory from the psychological and neurosciences perspectives: A review [Internet]. Vol. 9, Frontiers in Psychology. Frontiers Media S.A.; 2018. [cited 2021 Jan 8]. p. 401. Available from: www.frontiersin.org. doi: 10.3389/fpsyg.2018.00401 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Tine M. Working Memory Differences Between Children Living in Rural and Urban Poverty. J Cogn Dev [Internet]. 2014. Oct 2 [cited 2020 Nov 17];15(4):599–613. Available from: http://www.tandfonline.com/doi/abs/10.1080/15248372.2013.797906 [DOI] [PMC free article] [PubMed] [Google Scholar]
47.St. John AM, Kibbe M, Tarullo AR. A systematic assessment of socioeconomic status and executive functioning in early childhood. J Exp Child Psychol [Internet]. 2018;178:352–68. Available from: 10.1016/j.jecp.2018.09.003 [DOI] [PubMed] [Google Scholar]
48.Berry ED, Waterman AH, Baddeley AD, Hitch GJ, Allen RJ. The limits of visual working memory in children: Exploring prioritization and recency effects with sequential presentation. Dev Psychol [Internet]. 2018. [cited 2021 Jan 29];54(2). Available from: http://eprints.whiterose.ac.uk/120972/. doi: 10.1037/dev0000427 [DOI] [PubMed] [Google Scholar]
49.Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med [Internet]. 2009. Jul 21 [cited 2020 Dec 3];6(7):e1000097. Available from: https://dx.plos.org/10.1371/journal.pmed.1000097 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Pollock A, Berge E. How to do a systematic review [Internet]. Vol. 13, International Journal of Stroke. SAGE Publications Inc.; 2018. [cited 2020 Dec 1]. p. 138–56. Available from: http://journals.sagepub.com/doi/10.1177/1747493017743796 [DOI] [PubMed] [Google Scholar]
51.Prady SL, Uphoff EP, Power M, Golder S. Development and validation of a search filter to identify equity-focused studies: reducing the number needed to screen. BMC Med Res Methodol. 2018;18(1):1–9. doi: 10.1186/s12874-017-0458-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
52.O’Neill J, Tabish H, Welch V, Petticrew M, Pottie K, Clarke M, et al. Applying an equity lens to interventions: Using PROGRESS ensures consideration of socially stratifying factors to illuminate inequities in health. J Clin Epidemiol. 2014. Jan 1;67(1):56–64. doi: 10.1016/j.jclinepi.2013.08.005 [DOI] [PubMed] [Google Scholar]
53.Downes MJ, Brennan ML, Williams HC, Dean RS. Development of a critical appraisal tool to assess the quality of cross-sectional studies (AXIS). BMJ Open [Internet]. 2016. Dec 1 [cited 2020 Dec 1];6(12):e011458. Available from: http://bmjopen.bmj.com/. doi: 10.1136/bmjopen-2016-011458 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.NHBLI. Study Quality Assessment Tools | National Heart, Lung, and Blood Institue, National Institutes of Health [Internet]. 2017 [cited 2020 Dec 1]. Available from: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools.
55.Borenstein M, Hedges L V., Higgins JPT, Rothstein HR. Converting Among Effect Sizes. In: Introduction to Meta-analysis. John Wiley & Sons, Ltd; 2009. [Google Scholar]
56.Higgins JP, Green S. Cochrane handbook for systematic reviews of interventions (Vol. 4). John Wiley & Sons; 2011. [Google Scholar]
57.IntHout J, Ioannidis JPA, Rovers MM, Goeman JJ. Plea for routinely presenting prediction intervals in meta-analysis. BMJ Open [Internet]. 2016. Jul 12 [cited 2020 Dec 2];6(7):e010247. Available from: http://www.r-project.org/. doi: 10.1136/bmjopen-2015-010247 [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Lensing N, Elsner B. Development of hot and cool executive functions in middle childhood: Three-year growth curves of decision making and working memory updating. J Exp Child Psychol. 2018. Sep 1;173:187–204. doi: 10.1016/j.jecp.2018.04.002 [DOI] [PubMed] [Google Scholar]
59.Van den Noortgate W, López-López JA, Marín-Martínez F, Sánchez-Meca J. Meta-analysis of multiple outcomes: a multilevel approach. Behav Res Methods [Internet]. 2014. Nov 1 [cited 2020 Dec 2];47(4):1274–94. Available from: https://link.springer.com/article/10.3758/s13428-014-0527-2. [DOI] [PubMed] [Google Scholar]
60.Moeyaert M, Ugille M, Natasha Beretvas S, Ferron J, Bunuan R, Van den Noortgate W. Methods for dealing with multiple outcomes in meta-analysis: a comparison between averaging effect sizes, robust variance estimation and multilevel meta-analysis. Int J Soc Res Methodol [Internet]. 2017. Nov 2 [cited 2020 Dec 2];20(6):559–72. Available from: https://www.tandfonline.com/doi/full/10.1080/13645579.2016.1252189 [Google Scholar]
61.Hoyt WT, Del Re AC. Effect size calculation in meta-analyses of psychotherapy outcome research. Psychother Res [Internet]. 2018. May 4 [cited 2020 Dec 2];28(3):379–88. Available from: https://www.tandfonline.com/doi/abs/10.1080/10503307.2017.1405171 [DOI] [PubMed] [Google Scholar]
62.Tanner-Smith EE, Tipton E. Robust variance estimation with dependent effect sizes: Practical considerations including a software tutorial in Stata and SPSS. Res Synth Methods [Internet]. 2014. Mar 1 [cited 2020 Dec 2];5(1):13–30. Available from: https://onlinelibrary.wiley.com/doi/full/10.1002/jrsm.1091 [DOI] [PubMed] [Google Scholar]
63.Deeks JJ, Higgins JPT, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses | Cochrane Training [Internet]. 2021. [cited 2021 Jun 15]. Available from: https://training.cochrane.org/handbook/current/chapter-10#section-10-16. [Google Scholar]
64.Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. Br Med J [Internet]. 1997. Sep 13 [cited 2020 Dec 2];315(7109):629–34. Available from: https://www.bmj.com/content/315/7109/629 [DOI] [PMC free article] [PubMed] [Google Scholar]
65.StataCorp. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC. 2019. [Google Scholar]
66.Hedberg EC. ROBUMETA: Stata module to perform robust variance estimation in meta-regression with dependent effect size estimates [Internet]. Statistical Software Components. Boston College Department of Economics; 2014 [cited 2020 Nov 25]. Available from: https://ideas.repec.org/c/boc/bocode/s457219.html.
67.Ogilvie D, Fayter D, Petticrew M, Sowden A, Thomas S, Whitehead M, et al. The harvest plot: A method for synthesising evidence about the differential effects of interventions. BMC Med Res Methodol [Internet]. 2008. Feb 25 [cited 2020 Dec 2];8(1):1–7. Available from: https://link.springer.com/articles/doi: 10.1186/1471-2288-8-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Arán-Filippetti V. Structure and invariance of executive functioning tasks across socioeconomic status: Evidence from Spanish-speaking children. Span J Psychol. 2013;16:1–15. [Google Scholar]
69.Baddeley A, Hitch GJ, Allen RJ. A Multicomponent Model of Working Memory. In: Logie R, Camos V, & Cowan N (Eds), Working Memory: The State of the Science. Oxford University Press; 2021. [Google Scholar]
70.Engle RW, Kane MJ. Executive attention, working memory capacity, and a two-factor theory of cognitive control. In: Psychology of learning and motivation. 2004. p. 145–200. [Google Scholar]
71.Cowan N. What are the differences between long-term, short-term, and working memory? In: Progress in Brain Research. Elsevier; 2008. p. 323–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Deeks JJ, Higgins JP, Altman DG. Analysing data and undertaking meta-analyses. In: Cochrane Handbook for Systematic Reviews of Interventions [Internet]. Wiley; 2019. [cited 2020 Dec 1]. p. 241–84. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1002/9781119536604.ch10. [Google Scholar]
73.Borenstein M, Hedges L., Higgins J., Rothstein HR. Introduction to Meta-Analysis. John Wiley & Sons. 2011. [Google Scholar]
74.Rhoades BL, Greenberg MT, Lanza ST, Blair C. Demographic and familial predictors of early executive function development: Contribution of a person-centered perspective. J Exp Child Psychol [Internet]. 2011;108(3):638–62. Available from: doi: 10.1016/j.jecp.2010.08.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Evans GW, Schamberg MA. Childhood poverty, chronic stress, and adult working memory. Proc Natl Acad Sci U S A [Internet]. 2009. Apr 21 [cited 2020 Dec 2];106(16):6545–9. Available from: www.pnas.orgcgidoi10.1073pnas.0811910106 [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Goodman JB, Freeman EE, Chalmers KA. The relationship between early life stress and working memory in adulthood: A systematic review and meta-analysis. Memory. 2018;8211. doi: 10.1080/09658211.2018.1561897 [DOI] [PubMed] [Google Scholar]
77.Spencer B, Castano E. Social class is dead. Long live social class! Stereotype threat among low socioeconomic status individuals. Soc Justice Res [Internet]. 2007. Dec 7 [cited 2020 Dec 2];20(4):418–32. Available from: https://link.springer.com/article/10.1007/s11211-007-0047-7. [Google Scholar]
78.Appel M, Kronberger N. Stereotypes and the Achievement Gap: Stereotype Threat Prior to Test Taking. Educ Psychol Rev. 2012;24(4):609–35. [Google Scholar]
79.Schmader T, Johns M, Forbes C. An Integrated Process Model of Stereotype Threat Effects on Performance. Psychol Rev [Internet]. 2008. Apr [cited 2020 Dec 2];115(2):336–56. Available from: /record/2008-04236-003. doi: 10.1037/0033-295X.115.2.336 [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Désert M, Préaux M, Jund R. So young and already victims of stereotype threat: Socio-economic status and performance of 6 to 9 years old children on raven’s progressive matrices. Eur J Psychol Educ [Internet]. 2009. Jun [cited 2020 Dec 2];24(2):207–18. Available from: https://link.springer.com/article/10.1007/BF03173012. [Google Scholar]
81.Heberle AE, Carter AS. Cognitive aspects of young children’s experience of economic disadvantage. Psychol Bull [Internet]. 2015. Jul 1 [cited 2020 Dec 2];141(4):723–46. Available from: /record/2015-13568-001. doi: 10.1037/bul0000010 [DOI] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0260788.r001

Decision Letter 0

Muhammad Shahzad Aslam

7 Oct 2021

PONE-D-21-23197The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysisPLOS ONE

Dear,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by 1st Novermber 2021. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Muhammad Shahzad Aslam, Ph.D.,M.Phil., Pharm-D

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include in the Ethics statement in the Methods section the information that the IRB waived the need for informed consent from participants

3. We note that this manuscript is a systematic review or meta-analysis; our author guidelines therefore require that you use PRISMA guidance to help improve reporting quality of this type of study. Please upload copies of the completed PRISMA checklist as Supporting Information with a file name “PRISMA checklist

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Review for PLoS One

PONE-D-21-23197

The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysis

GENERAL COMMENTS: This is a very interesting review/meta-analysis, addressing a less-studied topic in the field of working memory – specifically, to examine the influence of socio-economic status on the development of working memory in children (magnitude and variability). The authors tackled this issue by reviewing publications of typically developing children from 0-18 years, with quantitative measures of working memory and SES, until June 2021. Meta-analyses were conducted with random effects analyses and Harvest Plots. The sample was very large, with 64 eligible studies and nearly 40,000 children aged 2 months to 18 years – the meta-analyses were conducted on half the sample. The authors found in the meta-analyses and supporting Harvest plot that reduced SES was linked to lower WM, particularly storage and information processing.

ABSTRACT COMMENTS:

• Comprehensive abstract, concise with relevant details and important conclusions about supporting the development of WM in children.

INTRODUCTION COMMENTS:

• Very clear introduction, framing the topic well, and highlighting the novelty/need for the current meta-analysis, good job!

• Minor point – page 6, from line 125 – repeated use of the word different makes sentence less fluent, advise revision.

• Good summary of aims to examine the components of working memory in line with children’s SES.

METHODS COMMENTS:

• Good use of the PEO framework

• Participants, exposure and outcomes all clearly described

• Clear study design

• Clear study inclusion criteria.

• Excellent search/selection strategy via PROGRESS

• Clear details about data extraction

• Excellent use of risk of bias tools

• Validity/reliability measures very thorough

RESULTS COMMENTS:

• Useful information about study characteristics, especially which countries

• Authors should be commended for the high number of individual participants

• Excellent summaries of heterogeneity, publication bias and sensitivity analyses

• The figures are excellent and help to illustrate the findings well.

• The summary paragraph at the end of the Harvest Plot section is very useful

• Overall the results section is excellent – clearly reported.

DISCUSSION COMMENTS:

• Excellent summary of findings – clearly interprets the wealth of data provided in the paper

• Good use of subheadings

• Good link between parents’ SES, allostatic load and variance in child working memory development

• Also the discussion about stereotype threat was intriguing

OVERALL RECOMMENDATION: An excellent paper, truly well executed and inspiring. Beyond the minor points above, this paper could be accepted in its current form.

• ACCEPT

Reviewer #2: Review of paper titled: The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysis

Journal: PLOS ONE

Introduction

The authors cover a substantive amount of relevant literature in the introduction to explain the possible mechanisms through which socioeconomic position may affect complex working memory in children under 18 years of age. The authors further highlight the disagreement and conflicting findings within the literature regarding the association between socioeconomic position and complex working memory; this is not justification enough for undertaking the current systematic review and meta-analysis. It would be good to see the authors explain why this information is clinically relevant and what the potential impact it could have for children currently living in lower socioeconomic positions.

Methodology

The authors state that their protocol is registered on PROSPERO, however, a quick search using the CRD number cited shows no hits. See screen shot below.

Why have the authors chosen the PEO framework as opposed to the PICOS framework which would include a comparison group/population? PICOS – population, intervention/exposure, comparison, outcome, study design. Based on the full description provided under the study eligibility criteria section, this review is able to meet the conditions for using the PICOS framework.

Although the PRISMA checklist does not specifically state that study selection needs to be done by two independent reviewers, the PRISMA statement website makes reference to the Cochrane Handbook under the protocol guidelines section. In the Cochrane handbook is explicitly states that all studies from the search should be reviewed by at least two people working independently to check if studies meet the inclusion criteria. See https://training.cochrane.org/handbook/archive/v6.1/chapter-04#section-4-6-4

For this study, the second reviewer only reviewed a subset of the excluded studies (not the included studies) as well as the data extraction. In light of this, how do the authors reasonably classify this study selection as being part of a standardized systematic review process?

A meta-analysis is typically a metric defined by two relationships, however, from the description of how they conducted the meta-analysis, it’s not explicitly clear that they had two groups to compare in this meta-analysis. Did the authors compare lower SEP to higher SEP in the meta-analysis? If so, I think it would help to explicitly state so. If not, they that also needs to be made clear and also provide a justification for taking the meta-analysis down to only one metric.

In the PRISMA flow diagram it shows that 8192 studies were obtained after duplicates were removed. From this a further 7881 studies were excluded, which would leave you with a total of 311 eligible studies and not 396 as indicated on the flow diagram.

Results

The result section is presented well. Figures 3 & 4 contain forest plots for which the scale -1 – 3 is provided. It make it simpler for the reader, please can the authors indicate which side of the midline favours the relationship in question.

Discussion

The discussion merely presents a summary of the findings and doesn’t really create an argument or provide an explanations about the statistical findings.

In the recommendations for future studies the authors state that “it has previously been found that socioeconomic disadvantage is associated with worse working memory in ethnic minority children, whilst ethnic majority children at different levels of socioeconomic risks have similar working memory ability”. Are the authors suggesting that ethnicity itself, inherently affects working memory despite socioeconomic position? Further more they state that the “disadvantage faced by ethnic minority groups may therefore exacerbate the negative association between socioeconomic position and working memory”; can the authors please elaborate on what these disadvantages are?

In the final sentence of the conclusion the authors reveal the impact of the findings of this review. The potential for this information to be clinically and practically useful should be mentioned much earlier in the manuscript (i.e.: as part of the justification for doing this in the introduction).

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Samantha J Brooks

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2021 Dec 2;16(12):e0260788. doi: 10.1371/journal.pone.0260788.r002

Author response to Decision Letter 0

27 Oct 2021

We thank the reviewers for their comments and think the revisions we have made in response have strengthened it considerably. Where required, our responses to the reviewers’ suggestions are given in bold after each statement. The page/line numbers relate to where the change has been made.

Journal

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Response: We have ensured that PLOS ONE’s style requirements are met.

2. Please include in the Ethics statement in the Methods section the information that the IRB waived the need for informed consent from participants

Response: Given this paper is a systematic review, informed consent was not applicable. For information on whether informed consent was sought for the individual studies included in the review, please see the individual studies themselves.

Response: The PRISMA checklist was uploaded with the original file submission. We have uploaded it again with this resubmission.

Reviewer #1

GENERAL COMMENTS:

This is a very interesting review/meta-analysis, addressing a less-studied topic in the field of working memory – specifically, to examine the influence of socio-economic status on the development of working memory in children (magnitude and variability). The authors tackled this issue by reviewing publications of typically developing children from 0-18 years, with quantitative measures of working memory and SES, until June 2021. Meta-analyses were conducted with random effects analyses and Harvest Plots. The sample was very large, with 64 eligible studies and nearly 40,000 children aged 2 months to 18 years – the meta-analyses were conducted on half the sample. The authors found in the meta-analyses and supporting Harvest plot that reduced SES was linked to lower WM, particularly storage and information processing.

ABSTRACT COMMENTS:

• Comprehensive abstract, concise with relevant details and important conclusions about supporting the development of WM in children.

INTRODUCTION COMMENTS:

• Very clear introduction, framing the topic well, and highlighting the novelty/need for the current meta-analysis, good job!

• Minor point – page 6, from line 125 – repeated use of the word different makes sentence less fluent, advise revision.

• Good summary of aims to examine the components of working memory in line with children’s SES.

Response: We appreciate that the repeated use of the word ‘different’ had reduced fluency in this section. We have now addressed this by amending our language (p.7, line 134-137)

METHODS COMMENTS:

• Good use of the PEO framework

• Participants, exposure and outcomes all clearly described

• Clear study design

• Clear study inclusion criteria.

• Excellent search/selection strategy via PROGRESS

• Clear details about data extraction

• Excellent use of risk of bias tools

• Validity/reliability measures very thorough

RESULTS COMMENTS:

• Useful information about study characteristics, especially which countries

• Authors should be commended for the high number of individual participants

• Excellent summaries of heterogeneity, publication bias and sensitivity analyses

• The figures are excellent and help to illustrate the findings well.

• The summary paragraph at the end of the Harvest Plot section is very useful

• Overall the results section is excellent – clearly reported.

DISCUSSION COMMENTS:

• Excellent summary of findings – clearly interprets the wealth of data provided in the paper

• Good use of subheadings

• Good link between parents’ SES, allostatic load and variance in child working memory development

• Also the discussion about stereotype threat was intriguing

OVERALL RECOMMENDATION:

• An excellent paper, truly well executed and inspiring. Beyond the minor points above, this paper could be accepted in its current form.

• ACCEPT

Response: Thank you very much for your thorough review that highlights the strengths of our paper. This was a very welcome review and we are very grateful to the Reviewer.

Reviewer #2

Introduction

Response: We respectfully disagree with the reviewer that the conflicting findings in the extant literature are not sufficient reason, without regard to other issues, to undertake a systematic review and meta-analysis. There is a contribution to be made to our knowledge (IS lower socioeconomic position associated with WM?) irrespective of the size of the effect or impact on the population. A very small attributable risk can have a large population impact if the risk factor is widespread, which is the case for low socioeconomic position.

We have, however, added a rationale for why clarifying this association is helpful in understanding inequalities in educational attainment to both the introduction and the discussion. We previously included some discussion of this in the introduction, where we explained how critical working memory is for successful learning and therefore for education (p.4 lines 60-68). This, in turn, means it is important to attempt to resolve the dispute in the literature around whether SEP impacts working memory (p. 4, line 70-74). In order to make this more explicit we have added some additional text in the introduction (p.4, line 67-68; p.5, line 80-86; p.7, lines 143-150) and discussion (p. 26, lines, 566-571 and p.29, line 649-652).

Methodology

The authors state that their protocol is registered on PROSPERO, however, a quick search using the CRD number cited shows no hits. See screen shot below.

Response: We have now included the full search term for the number, which should have included ‘CRD’ at the start (p.9, line 162)

Response: Whilst we acknowledge the merits of the PICOS framework for a systematic review of interventions, we feel that the PEO framework is a more appropriate framework for our review, as it fully captures all of our concepts of interest. If we had applied the PICOS framework, we would have repeated the criterion of ‘socioeconomic position’ for both exposure and comparison. We also made efforts to describe our eligible study designs following description of the PEO framework (p.9, line 177-183). We also note that this comment contrasts with Reviewer #1’s opinion, as their view was that the PEO framework had been used appropriately.

See https://training.cochrane.org/handbook/archive/v6.1/chapter-04#section-4-6-4

Response: We accept that it is ideal for all studies in a review to be screened by both reviewers, however, this is often not possible and is not an essential criteria according to Cochrane. The latest Cochrane handbook states that duplicate title and abstract screening is ideal, but that “it is acceptable that this initial screening of titles and abstracts is undertaken by only one person”. Our study fulfils these criteria as a subset of the excluded studies were reviewed, and therefore goes beyond screening by just one person.

See here: https://training.cochrane.org/handbook/current/chapter-04#section-4-6

Further, whilst we accept that the Cochrane handbook describes best practice for ideal reviewing processes – not all systematic reviews can feasibly follow all of the Cochrane guidelines, and are therefore not considered to be Cochrane reviews. However, such studies are still considered systematic reviews.

See here: https://www.cochranelibrary.com/about/about-cochrane-reviews

Response: We did compare lower SEP to higher SEP groups in the meta-analysis, and have amended the manuscript to make this clearer (p.12, line 255-257). This has also been made clearer by the below suggestion about the meta-analysis figures (p.17, line 367-368, p.18, line 380-381).

Response: Thank you for picking up on this error in our PRISMA diagram. This error occurred when we updated our search to include studies from a further two years, as we made a mistake when updating the PRISMA diagram to include the numbers from the new search. We have now amended the diagram to show that 8194 studies were screened, and 7798 of these were excluded at the title and abstract stage. The number of full text articles assessed for eligibility was 396. The new diagram has been uploaded with this resubmission.

Results

Response: We have amended the note beneath the figure to explain that the right hand side of the midline favours the higher socioeconomic groups (p.17, line 367-368, p.18, line 380-381).

Discussion

The discussion merely presents a summary of the findings and doesn’t really create an argument or provide an explanations about the statistical findings.

Response: We respectfully disagree that explanations about statistical findings are not provided in our discussion, and note that Reviewer 1’s opinion on this was that we “clearly interpret the wealth of data” in our discussion.

Potential explanations of our statistical findings are explored in the initial section of the discussion (see p.23, line 501-504; p.24 line 513 to 516 and line 521-525; p.25, line 532-537 and line 542 to 546). As noted by Reviewer 1, the ‘Directions for future research’ section also highlights possible explanations for the findings by considering mediating factors (p.27, line 596-604), and stereotype threat (p. 26, line 572-588).

Note also, that in response to the Reviewer 2’s earlier point, we have added to the discussion of the impact of our findings for children’s educational attainment, creating a stronger inference for the importance of our findings (p. 26, lines, 566-571 and p.29, line 650-652).

Response: We are not suggesting anything about causal pathways in making this statement, simply pointing to the associations that are present in the extant literature. Our review highlights that there is not yet sufficient evidence to enable any clarity about whether or not variation in working memory among different ethnic groups cannot be explained entirely by socioeconomic position. The study we cite finds a difference in social gradients for working memory across ethnic groups, and we have clarified this in the revised manuscript (p.26, line 578 to 588).

Response: Thank you for highlighting this. As stated in response to the Reviewer’s earlier comment, we have revised both our introduction (p.4, line 67-68; p.5, line 80-86; p.7, lines 143-150) and discussion (p. 26, lines, 566-571 and p.29, line 650-652) to emphasise this.

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(25.6KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0260788.r003

Decision Letter 1

Muhammad Shahzad Aslam

17 Nov 2021

The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysis

PONE-D-21-23197R1

Dear,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Muhammad Shahzad Aslam, Ph.D.,M.Phil., Pharm-D

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Reviewer #1: I am happy with the reviewer's response - it is a well-written paper that deserves publication - good job!

Reviewer #3: I agree with the first reviewer, as I believe the article is very well written, relevant and has a very suitable methodology. The results are well presented and well discussed.

The authors adequately responded to all review suggestions, making this version more complete.

I consider the article to be an asset to the state of the art.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Dr Samantha J Brooks

Reviewer #3: Yes: Sónia Quintão

PLoS One. doi: 10.1371/journal.pone.0260788.r004

Acceptance letter

Muhammad Shahzad Aslam

23 Nov 2021

PONE-D-21-23197R1

The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysis

Dear Dr. Mooney:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Muhammad Shahzad Aslam

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 File. Search strategy for Embase.

(DOCX)

Click here for additional data file.^{(25.3KB, docx)}

S2 File. Table of study characteristics.

(DOCX)

Click here for additional data file.^{(41.9KB, docx)}

S3 File. References for studies included in systematic review.

(DOCX)

Click here for additional data file.^{(28.6KB, docx)}

S4 File. PRISMA 2009 checklist.

(DOC)

Click here for additional data file.^{(63.5KB, doc)}

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(25.6KB, docx)}

Data Availability Statement

All relevant data are within the manuscript and its supporting information files.

[pone.0260788.ref001] 1.Baddeley A. Working memory. Curr Biol. 2010. [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref002] 2.Cowan N. The many faces of working memory and short-term storage. Psychon Bull Rev. 2017;24(4):1158–70. doi: 10.3758/s13423-016-1191-6 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref003] 3.Anderson P. Assessment and Development of Executive Function (EF) During Childhood. Child Neuropsychol [Internet]. 2002;8(2):71–82. Available from: http://www.tandfonline.com/doi/abs/10.1076/chin.8.2.71.8724 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref004] 4.Diamond A, Lee K. Interventions shown to aid executive function development in children. Science (80-). 2011;333(October):959–64. doi: 10.1126/science.1204529 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref005] 5.Gathercole SE, Durling E, Evans M, Jeffcock S, Stone S. Working memory abilities and children’s performance in laboratory analogues of classroom activities. Appl Cogn Psychol [Internet]. 2008. Dec 1 [cited 2021 Jan 4];22(8):1019–37. Available from: http://doi.wiley.com/10.1002/acp.1407. [Google Scholar]

[pone.0260788.ref006] 6.Waterman AH, Atkinson AL, Aslam SS, Holmes J, Jaroslawska A, Allen RJ. Do actions speak louder than words? Examining children’s ability to follow instructions. Mem Cogn [Internet]. 2017. Aug 1 [cited 2020 Dec 2];45(6):877–90. Available from: https://link.springer.com/article/10.3758/s13421-017-0702-7. doi: 10.3758/s13421-017-0702-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref007] 7.Allen RJ, Waterman AH. How does enactment affect the ability to follow instructions in working memory? Mem Cogn [Internet]. 2015. Apr 1 [cited 2020 Dec 2];43(3):555–61. Available from: https://link.springer.com/article/10.3758/s13421-014-0481-3 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref008] 8.Gathercole S, Alloway T. Working memory and learning: A practical guide for teachers. Sage; 2008. [Google Scholar]

[pone.0260788.ref009] 9.Cowan N. Working Memory Underpins Cognitive Development, Learning, and Education. 2014;197–223. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref010] 10.Alloway TP, Copello E. Working memory: The what, the why, and the how. Aust Educ Dev Psychol. 2013;30(2):105–18. [Google Scholar]

[pone.0260788.ref011] 11.Gathercole SE, Pickering SJ, Knight C, Stegmann Z. Working memory skills and educational attainment: Evidence from national curriculum assessments at 7 and 14 years of age. Appl Cogn Psychol. 2004;18(1):1–16. [Google Scholar]

[pone.0260788.ref012] 12.Holmes J, Adams JW. Educational Psychology Working Memory and Children’s Mathematical Skills: Implications for mathematical development and mathematics curricula. 2007. [cited 2020 Nov 11]; Available from: https://www.tandfonline.com/action/journalInformation?journalCode=cedp20. [Google Scholar]

[pone.0260788.ref013] 13.St Clair-Thompson HL, Gathercole SE. Executive functions and achievements in school: Shifting, updating, inhibition, and working memory. Q J Exp Psychol. 2006;59(4):745–59. [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref014] 14.Allen K, Higgins S, Adams J. The Relationship between Visuospatial Working Memory and Mathematical Performance in School-Aged Children: a Systematic Review. Educ Psychol Rev. 2019. doi: 10.1007/s10648-018-9458-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref015] 15.Peng P, Namkung J, Barnes M, Sun C. A meta-analysis of mathematics and working memory: Moderating effects of working memory domain, type of mathematics skill, and sample characteristics. J Educ Psychol [Internet]. 2016. May 1 [cited 2020 Nov 11];108(4):455–73. Available from: /record/2015-41049-001. [Google Scholar]

[pone.0260788.ref016] 16.Peng P, Barnes M, Wang C, Swanson HL. A Meta-Analysis on the Relation Between Reading and Working Memory. 2017. [cited 2020 Nov 11]; Available from: 10.1037/bul0000124.supp. [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref017] 17.Carretti B, Borella E, Cornoldi C, De Beni R. Role of working memory in explaining the performance of individuals with specific reading comprehension difficulties: A meta-analysis. Learn Individ Differ [Internet]. 2009;19(2):246–51. Available from: 10.1016/j.lindif.2008.10.002. [DOI] [Google Scholar]

[pone.0260788.ref018] 18.Gruszka A, Nęcka E. Limitations of working memory capacity: The cognitive and social consequences. Eur Manag J. 2017. Dec 1;35(6):776–84. [Google Scholar]

[pone.0260788.ref019] 19.Krieger N, Williams DR, Moss NE. Measuring Social Class in US Public Health Research: Concepts, Methodologies, and Guidelines. Annu Rev Public Health [Internet]. 1997;18(1):341–78. Available from: http://www.annualreviews.org/doi/10.1146/annurev.publhealth.18.1.341. [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref020] 20.Rees G, Gromada A, Cuesta J, Bruckauf Z. An Unfair Start. Inequality in Children’s Education in Rich Countries. 2018. [Google Scholar]

[pone.0260788.ref021] 21.Halle T, Forry N, Hair E, Perper K, Wandner L, Wessel J, et al. Disparities in Early Learning and Development: Lessons from the Early Childhood Longitudinal Study-Birth Cohort (ECLS-B) [Internet]. 2009. [cited 2020 Dec 2]. Available from: http://www.nccp.org/publications/pdf/text_851.pdf. [Google Scholar]

[pone.0260788.ref022] 22.Feinstein L (London S of E. Inequality in the Early Cognitive Development of British Children in the 1970 Cohort. 2003;73–97. [Google Scholar]

[pone.0260788.ref023] 23.Jerrim J, Vignoles A. Social mobility, regression to the mean and the cognitive development of high ability children from disadvantaged homes. J R Stat Soc Ser A Stat Soc. 2013;176(4):887–906. [Google Scholar]

[pone.0260788.ref024] 24.Chattoo S, Atkin K. Understanding ‘race’ and ethnicity 2e: Theory, history, policy, practice. 2nd ed. Policy Press; 2019. doi: 10.1242/jcs.224147 [DOI] [Google Scholar]

[pone.0260788.ref025] 25.Ethnicity Parsons C., gender, deprivation and low educational attainment in England: Political arithmetic, ideological stances and the deficient society. Educ Citizsh Soc Justice [Internet]. 2016. Jul 8 [cited 2020 Nov 25];11(2):160–83. Available from: http://journals.sagepub.com/doi/10.1177/1746197916648282. [Google Scholar]

[pone.0260788.ref026] 26.Bornstein MH. Toward a model of culture↔parent↔child transactions. In: The transactional model of development: How children and contexts shape each other [Internet]. American Psychological Association; 2009. [cited 2020 Nov 25]. p. 139–61. Available from: /record/2009-05801-008. [Google Scholar]

[pone.0260788.ref027] 27.Duncan GJ, Magnuson K, Votruba-Drzal E. Moving beyond Correlations in Assessing the Consequences of Poverty. Annu Rev Psychol. 2017;68:413–34. doi: 10.1146/annurev-psych-010416-044224 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref028] 28.Lupien SJ, King S, Meaney MJ, McEwen BS. Can poverty get under your skin? Basal cortisol levels and cognitive function in children from low and high socioeconomic status. Dev Psychopathol [Internet]. 2001. Jun [cited 2020 Dec 4];13(3):653–76. Available from: https://www.cambridge.org/core/journals/development-and-psychopathology/article/abs/can-poverty-get-under-your-skin-basal-cortisol-levels-and-cognitive-function-in-children-from-low-and-high-socioeconomic-status/44990AC6E1E448F8414FD171FE25CA5D. doi: 10.1017/s0954579401003133 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref029] 29.Votruba-Drzal E. Economic disparities in middle childhood development: Does income matter? Dev Psychol. 2006. Nov;42(6):1154–67. doi: 10.1037/0012-1649.42.6.1154 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref030] 30.Amso D, Lynn A. Distinctive Mechanisms of Adversity and Socioeconomic Inequality in Child Development: A Review and Recommendations for Evidence-Based Policy. Policy Insights from Behav Brain Sci. 2017;4(2):139–46. doi: 10.1177/2372732217721933 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref031] 31.Tucker-Drob EM, Briley DA, Harden KP. Genetic and Environmental Influences on Cognition Across Development and Context. Curr Dir Psychol Sci [Internet]. 2013. Oct 25 [cited 2021 Apr 7];22(5):349–55. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24799770 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref032] 32.Peng P, Kievit RA. The Development of Academic Achievement and Cognitive Abilities: A Bidirectional Perspective. Child Dev Perspect [Internet]. 2020. Mar 27 [cited 2021 Apr 7];14(1):15–20. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1111/cdep.12352. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref033] 33.Engel PMJ, Santos FH, Gathercole SE. Are working memory measures free of socio-economic influence? Anim Genet. 2008;39(5):561–3. doi: 10.1111/j.1365-2052.2008.01757.x [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref034] 34.Vandenbroucke L, Verschueren K, Ceulemans E, De Smedt B, De Roover K, Baeyens D. Family demographic profiles and their relationship with the quality of executive functioning subcomponents in kindergarten. Br J Dev Psychol. 2016. Jun 1;34(2):226–44. doi: 10.1111/bjdp.12127 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref035] 35.Wiebe SA, Espy KA, Charak D. Using Confirmatory Factor Analysis to Understand Executive Control in Preschool Children: I. Latent Structure. Dev Psychol. 2008;44(2):575–87. doi: 10.1037/0012-1649.44.2.575 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref036] 36.Hackman DA, Betancourt LM, Gallop R, Romer D, Brodsky NL, Hurt H, et al. Mapping the trajectory of socioeconomic disparity in working memory: Parental and neighborhood factors. Child Dev. 2014;85(4):1433–45. doi: 10.1111/cdev.12242 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref037] 37.Lawson GM, Farah MJ. Executive function as a mediator between SES and academic achievement throughout childhood. Int J Behav Dev. 2017;41(1):94–104. doi: 10.1177/0165025415603489 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref038] 38.Farah MJ, Shera DM, Savage JH, Betancourt L, Giannetta JM, Brodsky NL, et al. Childhood poverty: Specific associations with neurocognitive development. Brain Res. 2006;1110(1):166–74. doi: 10.1016/j.brainres.2006.06.072 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref039] 39.Lawson GM, Hook CJ, Farah MJ. A meta-analysis of the relationship between socioeconomic status and executive function performance among children. Dev Sci. 2018;21(2):1–22. doi: 10.1111/desc.12529 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref040] 40.Baddeley AD. Working memory: Theories, models, and controversies. Explor Work Mem Sel Work Alan Baddeley. 2012;333–69. [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref041] 41.Cowan N, Morey CC, Naveh-Benjamin M. An Embedded-Processes Approach to Working Memory. In: Logie R, Camos V, & Cowan N (Eds), Working Memory: The State of the Science. Oxford University Press.; 2021. [Google Scholar]

[pone.0260788.ref042] 42.Oberauer K. Towards a theory of working memory: From metaphors to mechanisms. In: Logie R, Camos V, & Cowan N (Eds), Working Memory: The State of the Science. Oxford University Press; 2021. [Google Scholar]

[pone.0260788.ref043] 43.Pickering S, Gathercole SE. Working memory test battery for children (WMTB-C). Psychological Corporation; 2001. [Google Scholar]

[pone.0260788.ref044] 44.D’Esposito M, Postle BR. The cognitive neuroscience of working memory. Annu Rev Psychol [Internet]. 2015. Jan 1 [cited 2021 Jan 8];66:115–42. Available from: www.annualreviews.org. doi: 10.1146/annurev-psych-010814-015031 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref045] 45.Chai WJ, Abd Hamid AI, Abdullah JM. Working memory from the psychological and neurosciences perspectives: A review [Internet]. Vol. 9, Frontiers in Psychology. Frontiers Media S.A.; 2018. [cited 2021 Jan 8]. p. 401. Available from: www.frontiersin.org. doi: 10.3389/fpsyg.2018.00401 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref046] 46.Tine M. Working Memory Differences Between Children Living in Rural and Urban Poverty. J Cogn Dev [Internet]. 2014. Oct 2 [cited 2020 Nov 17];15(4):599–613. Available from: http://www.tandfonline.com/doi/abs/10.1080/15248372.2013.797906 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref047] 47.St. John AM, Kibbe M, Tarullo AR. A systematic assessment of socioeconomic status and executive functioning in early childhood. J Exp Child Psychol [Internet]. 2018;178:352–68. Available from: 10.1016/j.jecp.2018.09.003 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref048] 48.Berry ED, Waterman AH, Baddeley AD, Hitch GJ, Allen RJ. The limits of visual working memory in children: Exploring prioritization and recency effects with sequential presentation. Dev Psychol [Internet]. 2018. [cited 2021 Jan 29];54(2). Available from: http://eprints.whiterose.ac.uk/120972/. doi: 10.1037/dev0000427 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref049] 49.Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med [Internet]. 2009. Jul 21 [cited 2020 Dec 3];6(7):e1000097. Available from: https://dx.plos.org/10.1371/journal.pmed.1000097 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref050] 50.Pollock A, Berge E. How to do a systematic review [Internet]. Vol. 13, International Journal of Stroke. SAGE Publications Inc.; 2018. [cited 2020 Dec 1]. p. 138–56. Available from: http://journals.sagepub.com/doi/10.1177/1747493017743796 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref051] 51.Prady SL, Uphoff EP, Power M, Golder S. Development and validation of a search filter to identify equity-focused studies: reducing the number needed to screen. BMC Med Res Methodol. 2018;18(1):1–9. doi: 10.1186/s12874-017-0458-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref052] 52.O’Neill J, Tabish H, Welch V, Petticrew M, Pottie K, Clarke M, et al. Applying an equity lens to interventions: Using PROGRESS ensures consideration of socially stratifying factors to illuminate inequities in health. J Clin Epidemiol. 2014. Jan 1;67(1):56–64. doi: 10.1016/j.jclinepi.2013.08.005 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref053] 53.Downes MJ, Brennan ML, Williams HC, Dean RS. Development of a critical appraisal tool to assess the quality of cross-sectional studies (AXIS). BMJ Open [Internet]. 2016. Dec 1 [cited 2020 Dec 1];6(12):e011458. Available from: http://bmjopen.bmj.com/. doi: 10.1136/bmjopen-2016-011458 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref054] 54.NHBLI. Study Quality Assessment Tools | National Heart, Lung, and Blood Institue, National Institutes of Health [Internet]. 2017 [cited 2020 Dec 1]. Available from: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools.

[pone.0260788.ref055] 55.Borenstein M, Hedges L V., Higgins JPT, Rothstein HR. Converting Among Effect Sizes. In: Introduction to Meta-analysis. John Wiley & Sons, Ltd; 2009. [Google Scholar]

[pone.0260788.ref056] 56.Higgins JP, Green S. Cochrane handbook for systematic reviews of interventions (Vol. 4). John Wiley & Sons; 2011. [Google Scholar]

[pone.0260788.ref057] 57.IntHout J, Ioannidis JPA, Rovers MM, Goeman JJ. Plea for routinely presenting prediction intervals in meta-analysis. BMJ Open [Internet]. 2016. Jul 12 [cited 2020 Dec 2];6(7):e010247. Available from: http://www.r-project.org/. doi: 10.1136/bmjopen-2015-010247 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref058] 58.Lensing N, Elsner B. Development of hot and cool executive functions in middle childhood: Three-year growth curves of decision making and working memory updating. J Exp Child Psychol. 2018. Sep 1;173:187–204. doi: 10.1016/j.jecp.2018.04.002 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref059] 59.Van den Noortgate W, López-López JA, Marín-Martínez F, Sánchez-Meca J. Meta-analysis of multiple outcomes: a multilevel approach. Behav Res Methods [Internet]. 2014. Nov 1 [cited 2020 Dec 2];47(4):1274–94. Available from: https://link.springer.com/article/10.3758/s13428-014-0527-2. [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref060] 60.Moeyaert M, Ugille M, Natasha Beretvas S, Ferron J, Bunuan R, Van den Noortgate W. Methods for dealing with multiple outcomes in meta-analysis: a comparison between averaging effect sizes, robust variance estimation and multilevel meta-analysis. Int J Soc Res Methodol [Internet]. 2017. Nov 2 [cited 2020 Dec 2];20(6):559–72. Available from: https://www.tandfonline.com/doi/full/10.1080/13645579.2016.1252189 [Google Scholar]

[pone.0260788.ref061] 61.Hoyt WT, Del Re AC. Effect size calculation in meta-analyses of psychotherapy outcome research. Psychother Res [Internet]. 2018. May 4 [cited 2020 Dec 2];28(3):379–88. Available from: https://www.tandfonline.com/doi/abs/10.1080/10503307.2017.1405171 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref062] 62.Tanner-Smith EE, Tipton E. Robust variance estimation with dependent effect sizes: Practical considerations including a software tutorial in Stata and SPSS. Res Synth Methods [Internet]. 2014. Mar 1 [cited 2020 Dec 2];5(1):13–30. Available from: https://onlinelibrary.wiley.com/doi/full/10.1002/jrsm.1091 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref063] 63.Deeks JJ, Higgins JPT, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses | Cochrane Training [Internet]. 2021. [cited 2021 Jun 15]. Available from: https://training.cochrane.org/handbook/current/chapter-10#section-10-16. [Google Scholar]

[pone.0260788.ref064] 64.Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. Br Med J [Internet]. 1997. Sep 13 [cited 2020 Dec 2];315(7109):629–34. Available from: https://www.bmj.com/content/315/7109/629 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref065] 65.StataCorp. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC. 2019. [Google Scholar]

[pone.0260788.ref066] 66.Hedberg EC. ROBUMETA: Stata module to perform robust variance estimation in meta-regression with dependent effect size estimates [Internet]. Statistical Software Components. Boston College Department of Economics; 2014 [cited 2020 Nov 25]. Available from: https://ideas.repec.org/c/boc/bocode/s457219.html.

[pone.0260788.ref067] 67.Ogilvie D, Fayter D, Petticrew M, Sowden A, Thomas S, Whitehead M, et al. The harvest plot: A method for synthesising evidence about the differential effects of interventions. BMC Med Res Methodol [Internet]. 2008. Feb 25 [cited 2020 Dec 2];8(1):1–7. Available from: https://link.springer.com/articles/doi: 10.1186/1471-2288-8-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref068] 68.Arán-Filippetti V. Structure and invariance of executive functioning tasks across socioeconomic status: Evidence from Spanish-speaking children. Span J Psychol. 2013;16:1–15. [Google Scholar]

[pone.0260788.ref069] 69.Baddeley A, Hitch GJ, Allen RJ. A Multicomponent Model of Working Memory. In: Logie R, Camos V, & Cowan N (Eds), Working Memory: The State of the Science. Oxford University Press; 2021. [Google Scholar]

[pone.0260788.ref070] 70.Engle RW, Kane MJ. Executive attention, working memory capacity, and a two-factor theory of cognitive control. In: Psychology of learning and motivation. 2004. p. 145–200. [Google Scholar]

[pone.0260788.ref071] 71.Cowan N. What are the differences between long-term, short-term, and working memory? In: Progress in Brain Research. Elsevier; 2008. p. 323–38. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref072] 72.Deeks JJ, Higgins JP, Altman DG. Analysing data and undertaking meta-analyses. In: Cochrane Handbook for Systematic Reviews of Interventions [Internet]. Wiley; 2019. [cited 2020 Dec 1]. p. 241–84. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1002/9781119536604.ch10. [Google Scholar]

[pone.0260788.ref073] 73.Borenstein M, Hedges L., Higgins J., Rothstein HR. Introduction to Meta-Analysis. John Wiley & Sons. 2011. [Google Scholar]

[pone.0260788.ref074] 74.Rhoades BL, Greenberg MT, Lanza ST, Blair C. Demographic and familial predictors of early executive function development: Contribution of a person-centered perspective. J Exp Child Psychol [Internet]. 2011;108(3):638–62. Available from: doi: 10.1016/j.jecp.2010.08.004 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref075] 75.Evans GW, Schamberg MA. Childhood poverty, chronic stress, and adult working memory. Proc Natl Acad Sci U S A [Internet]. 2009. Apr 21 [cited 2020 Dec 2];106(16):6545–9. Available from: www.pnas.orgcgidoi10.1073pnas.0811910106 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref076] 76.Goodman JB, Freeman EE, Chalmers KA. The relationship between early life stress and working memory in adulthood: A systematic review and meta-analysis. Memory. 2018;8211. doi: 10.1080/09658211.2018.1561897 [DOI] [PubMed] [Google Scholar]

[pone.0260788.ref077] 77.Spencer B, Castano E. Social class is dead. Long live social class! Stereotype threat among low socioeconomic status individuals. Soc Justice Res [Internet]. 2007. Dec 7 [cited 2020 Dec 2];20(4):418–32. Available from: https://link.springer.com/article/10.1007/s11211-007-0047-7. [Google Scholar]

[pone.0260788.ref078] 78.Appel M, Kronberger N. Stereotypes and the Achievement Gap: Stereotype Threat Prior to Test Taking. Educ Psychol Rev. 2012;24(4):609–35. [Google Scholar]

[pone.0260788.ref079] 79.Schmader T, Johns M, Forbes C. An Integrated Process Model of Stereotype Threat Effects on Performance. Psychol Rev [Internet]. 2008. Apr [cited 2020 Dec 2];115(2):336–56. Available from: /record/2008-04236-003. doi: 10.1037/0033-295X.115.2.336 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0260788.ref080] 80.Désert M, Préaux M, Jund R. So young and already victims of stereotype threat: Socio-economic status and performance of 6 to 9 years old children on raven’s progressive matrices. Eur J Psychol Educ [Internet]. 2009. Jun [cited 2020 Dec 2];24(2):207–18. Available from: https://link.springer.com/article/10.1007/BF03173012. [Google Scholar]

[pone.0260788.ref081] 81.Heberle AE, Carter AS. Cognitive aspects of young children’s experience of economic disadvantage. Psychol Bull [Internet]. 2015. Jul 1 [cited 2020 Dec 2];141(4):723–46. Available from: /record/2015-13568-001. doi: 10.1037/bul0000010 [DOI] [PubMed] [Google Scholar]

PERMALINK

The association between socioeconomic disadvantage and children’s working memory abilities: A systematic review and meta-analysis

Kate E Mooney

Stephanie L Prady

Mary M Barker

Kate E Pickett

Amanda H Waterman

Roles

Abstract

Background and objective

Methods

Key findings

Conclusion

Introduction

Methods

Protocol, registration, and reporting standards

Study eligibility criteria

Population, exposure, and outcome

Study designs

Study inclusion criteria

Search and selection procedures

Data extraction

Risk of bias in individual studies

Validity and reliability of review

Data synthesis

Fig 1. Selection process and eligibility for inclusion in meta-analyses or Harvest plot.

Data synthesis: Meta-analytic methods

Investigation of heterogeneity

Sensitivity analyses

Meta regression analyses

Publication bias

Software

Data synthesis: Harvest plot

Results

Study selection

Fig 2. PRISMA 2009 flow diagram for all included studies.

Study characteristics

Meta-analyses

Summary of effects

Fig 3. Meta-analysis of the association between socioeconomic position and simple working memory (sorted by effect size).

Fig 4. Meta-analysis of the association between socioeconomic position and complex working memory (sorted by effect size).

Heterogeneity

Publication bias

Fig 5. Funnel plot for simple working memory meta-analyses.

Fig 6. Funnel plot for complex working memory meta-analyses.

Sensitivity analysis

Meta regression analyses

Table 1. Meta-regression analyses results.

Harvest plot

Fig 7. Harvest plot of the association between different socioeconomic position indicators with composite working memory, simple working memory, and complex working memory.

Discussion

Directions for future research

Strengths and limitations

Conclusion

Supporting information

Data Availability

Funding Statement

References

Decision Letter 0

Muhammad Shahzad Aslam

Roles

Author response to Decision Letter 0

Decision Letter 1

Muhammad Shahzad Aslam

Roles

Acceptance letter

Muhammad Shahzad Aslam

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases