Cognitive Stimulation as a Mechanism Linking Socioeconomic Status WithExecutive Function: A Longitudinal Investigation

Maya L. RosenUniversity of Washington and Harvard University

McKenzie P. HagenUniversity of Washington and Stanford University

Lucy A. LurieUniversity of Washington and Harvard University

Zoe E. MilesUniversity of Washington

Margaret A. SheridanUniversity of North Carolina, Chapel Hill

Andrew N. MeltzoffUniversity of Washington

Katie A. McLaughlinHarvard University

Executive functions (EF), including working memory, inhibition, and cognitive flexibility, vary as a functionof socioeconomic status (SES), with children from economically disadvantaged backgrounds having poorerperformance than their higher SES peers. Using observational methods, we investigated cognitive stimulationin the home as a mechanism linking SES with EF. In a sample of 101 children aged 60–75 months, cognitivestimulation fully mediated SES-related differences in EF. Critically, cognitive stimulation was positively associ-ated with the development of inhibition and cognitive flexibility across an 18-month follow-up period. Fur-thermore, EF at T1 explained SES-related differences in academic achievement at T2. Early cognitivestimulation—a modifiable factor—may be a desirable target for interventions designed to ameliorate SES-re-lated differences in cognitive development and academic achievement.

Executive functions (EF), including working mem-ory, inhibitory control, and cognitive flexibility, area set of cognitive skills involved in goal-directedbehavior that are critical for adaptive functioningand academic success (Blair & Razza, 2007;Samuels, Tournaki, Blackman, & Zilinski, 2016).Children’s EF ability varies as a function of socioe-conomic status (SES), such that children from fami-lies with greater income and parental educationperform better on EF tasks than children fromlower-SES families (Dilworth-Bart, 2012; Hackman

& Farah, 2009; Lawson, Hook, & Farah, 2018;Noble, McCandliss, & Farah, 2007; Noble, Norman,& Farah, 2005; Rosen, Sheridan, Sambrook, Melt-zoff, & McLaughlin, 2018). SES-related disparities inacademic achievement appear to be explained, atleast in part, by these differences in EF (Finn et al.,2016; Rosen et al., 2018). Although considerableresearch has been dedicated to understanding thisassociation, the underlying mechanisms explainingwhy SES is related to EF remain poorly understood.This study sought to evaluate whether the amountof cognitive stimulation in the home environment—measured using observational methods—is a mech-anism accounting for SES-related differences in EFperformance.

Prior research has observed a positive associationbetween family SES and EF ability in youth, suchthat children raised by more highly educated par-ents and in families with greater income perform

This work was supported by the National Institute of ChildHealth and Human Development at the National Institute ofHealth (F32 HD089514 to Maya L. Rosen), National Institute ofMental Health at the National Institutes of Health (R01-MH103291 and R01-MH106482 to Katie A. McLaughlin), theBrain and Behavior Foundation NARSAD Early InvestigatorAward, an Early Career Research Fellowship from the JacobsFoundation, and the IMHRO Rising Star Award to KM, and bysupport from the University of Washington I-LABS InnovativeResearch Fund.

Correspondence concerning this article should be addressed toMaya L. Rosen, Postdoctoral Research Fellow, 33 Kirkland St.,Cambridge, MA 02139. Electronic mail may be sent to mayalrosen@fas.harvard.edu.

© 2019 Society for Research in Child DevelopmentAll rights reserved. 0009-3920/2019/xxxx-xxxxDOI: 10.1111/cdev.13315

Child Development, xxxx 2019, Volume 00, Number 0, Pages 1–18

better on tests of working memory, inhibitory con-trol, and cognitive flexibility than children raised inlower-SES families (Farah et al., 2006). The positiveassociation between SES and EF is present in earlychildhood (Clearfield & Niman, 2012; Lipina, Mar-telli, Vuelta, & Colombo, 2005) and the gap neitherwidens nor narrows across development (Hackman,Gallop, Evans, & Farah, 2015). While some studieshave found that EF does not vary as a function ofSES (e.g., Engel, Santos, & Gathercole, 2008; Wiebe,Espy, & Charak, 2008), a recent meta-analysis thatincluded data from thousands of children aged 2 to18 years found a small-to-medium associationbetween SES and EF and a stronger associationamong studies with multiple measures of EF (Law-son et al., 2018).

Despite evidence for an association between SESand EF, the mechanisms that explain why SES isrelated to EF remain poorly understood. In particu-lar, the features of the early environment that varyas a function of SES and, in turn, may shape indi-vidual differences in EF, are largely unknown. Iden-tifying the specific aspects of early environmentalexperience that explain the association between SESand EF is critical to developing effective early inter-ventions to close this gap. A variety of potentialenvironmental mechanisms linking SES in child-hood and EF abilities have been proposed. Theseinclude: SES-related differences in parenting andinteractions with caregivers, environmental pre-dictability, exposure to toxins, poor nutrition, expo-sure to violence, and stress (Hackman & Farah,2009; Hackman, Farah, & Meaney, 2010; Johnson,Riis, & Noble, 2016). Many mechanistic modelsexplaining SES-related disparities in EF havefocused on aspects of early experience, such asenvironmental enrichment, parental scaffolding ofchild learning, parental warmth, and languageexposure (Carlson, 2009; Hackman et al., 2010; Len-gua et al., 2014; Sheridan & McLaughlin, 2014;Sheridan & McLaughlin, 2016). We have recentlyproposed an integrated mechanistic model in whichcognitive stimulation in the home environment—in-cluding parental involvement in learning, environ-mental complexity, and language quality andquantity—is a critical link explaining SES-relateddifferences in EF (Rosen, Amso, & McLaughlin,2019). Specifically, this model argues that interac-tion with caregivers early in development, coupledwith an environment rich with complex sensorystimuli, plays a central role in the development ofEF. In this model, cognitive stimulation encom-passes four domains of early experience, includingaccess to learning materials, caregiver involvement

in learning, variety of experiences, and the quantityand quality of linguistic experience. This modelproposes that cognitive stimulation is critical for EFdevelopment because caregivers guide attentionand promote associative learning through languageand other forms of social interaction that highlightfeatures of the environment that require children tosustain and regulate attention and to resolve con-flict between stimuli with overlapping features asthey build semantic representations of differentstimulus types (Rosen et al., 2019). These interac-tions provide critical scaffolding for the develop-ment of the prefrontal cortex and EF. Here, wedirectly test the hypothesis that cognitive stimula-tion is a mechanism explaining the associationbetween SES and EF—including working memory,inhibition, and cognitive flexibility.

To be a plausible environmental mechanismlinking SES with EF, cognitive stimulation mustvary with SES. Indeed, prior research has foundthat children raised in lower SES-environments areexposed to lower levels of cognitive stimulation(Bradley & Corwyn, 2002; Bradley, Corwyn, McA-doo, & Garcia Coll, 2001; Hackman et al., 2015;Hart & Risley, 1995; Lengua, Honorado, & Bush,2007; Lengua et al., 2014). In pioneering workusing observations of the home environment, Brad-ley and Corwyn (2002) found that children fromlow-SES backgrounds had reduced access to enrich-ing experiences, access to educational materialsincluding books, and more limited parentalinvolvement in learning such as teaching childrento read (possibly due to time demands or parentaleducation level). Other studies that have relied onparental reports have also found that parental edu-cation and family income are positively associatedwith the presence of cognitively stimulating materi-als and experiences (e.g., presence of books in thehouse), the degree of caregiver involvement in chil-dren’s learning, and access to enriching experiencesoutside of the home (Christensen, Schieve, Devine,& Drews-Botsch, 2014; Hackman et al., 2015; Rosenet al., 2018).

The quantity and quality of linguistic experiencesare another critical and well-studied aspect of cog-nitive stimulation that varies as a function of SES.In an early study, Hart and Risley found that thatchildren in lower SES households were exposed tosignificantly fewer words than their higher SEScounterparts (Hart & Risley, 1995). While this studywas small, recent work has replicated the findingthat SES is associated with the quantity of languageexposure in children in larger samples and usingtechnological advancements to more accurately

2 Rosen et al.

track language exposure in the home (Fernald,Marchman, & Weisleder, 2013; Gilkerson et al.,2017). Furthermore, language quality also varies asa function of SES, such that higher SES parents usegreater variety of words and more complex syntac-tical structure (Rowe, 2012). These specific, measur-able differences in language exposure have in turnbeen associated with disparities in child languageability and vocabulary (Fernald et al., 2013;Ram�ırez-Esparza, Garc�ıa-Sierra, & Kuhl, 2014;Romeo et al., 2018). A recent study also found thatmaternal language complexity and vocabularydiversity measured in early childhood in the labora-tory were associated with child EF later in develop-ment (Daneri, Blair, & Kuhn, 2018). Maternallanguage was associated with child vocabulary,which in turn mediated the association betweenmaternal language and child EF. Taken together,the above studies highlight that children reared inlower SES environments tend to experience lowerlevels of cognitive stimulation in the home includ-ing access to learning materials, caregiver involve-ment in learning, access to enriching experiences, aswell as reduced quantity and quality of linguisticexperiences.

Three studies have directly tested whether SES-related differences in EF are explained by differ-ences in cognitive stimulation. One study investi-gated the role of cognitive stimulation in EF abilityin 8–12 year olds and found significant associationsof access to learning materials and enrichmentactivities with EF and that enrichment activitiesmediated the association of SES with workingmemory and inhibition (Sarsour et al., 2011). Arecent study in children aged 7–17 used parentreport of enrichment activities and found that varia-tion in cognitive stimulation and enrichment pre-dicts working memory performance even at thehigh end of the SES distribution and mediates SES-related differences in working memory performance(Amso, Salhi, & Badre, 2018). However, the cross-sectional nature of these studies and the focus onolder children make it difficult to determinewhether cognitive stimulation plays a role in thelink between SES and the development of EF overtime. Indeed, EF disparities as a function of SESemerge quite early in development (Clearfield &Niman, 2012; Lipina et al., 2005). The only longitu-dinal study examining this question relied on par-ent-report to assess how aspects of the homeenvironment might explain SES-related differencesin EF along with a wide range of other potentialmediators (Hackman et al., 2015). That study foundthat enrichment—a composite score that included

access to learning materials, variety of experiences,and parental involvement in learning measuredrepeatedly across infancy and early childhood from6 to 54 months—mediated the association betweenSES and both working memory and planning at54 months, whereas other mechanisms (parentalstress, negative life events, maternal depression,and birth weight), did not explain this association.While this study provides support for the idea thatSES-related differences in EF can be explained, atleast in part, by cognitive stimulation, assessmentsrelied on parent-report of cognitive stimulationrather than in-home observations, used a limitedset of EF measures, and did not examine EF growthover time. Here, we examine the role of cognitivestimulation in the home environment, assessedusing observational methods, as a potential mecha-nism underlying the longitudinal associationbetween SES and the development of EF abilitiesacross the domains of working memory, inhibition,and cognitive flexibility.

The present study investigated the hypothesisthat cognitive stimulation—including access tolearning materials, parental involvement in learn-ing, and language exposure—is a mechanismexplaining the association between SES and EF. Weassessed cognitive stimulation in the homes of 60-to 75-month-old children from a wide range of SESbackgrounds using observational and structuredinterview metrics to quantify learning materials,caregiver involvement in learning, and variety ofexperiences, as well as a naturalistic assessment oflanguage quantity and quality in the home. Chil-dren performed tasks to test the three majordomains of EF: working memory, inhibition, andcognitive flexibility (Miyake, Friedman, Rettinger,Shah, & Hegarty, 2001) which are particularlyimportant for school readiness and academicachievement (Blair, 2002; Coldren, 2013; Finn et al.,2016). One to 2-years later, children came into thelaboratory and performed the same EF tasks as wellas tests of academic achievement. We hypothesizedthat cognitive stimulation would be a mechanismexplaining SES-related differences in EF concur-rently as well as account for growth in EF overtime and that SES-related differences in academicachievement would be explained by EF. Childrengrowing up in low-SES environments often experi-ence other adverse environmental experiences,including exposure to violence and even maltreat-ment (McLaughlin et al., 2012). Some have arguedthat exposure to violence may impact the develop-ment of EF (e.g., Hanson et al., 2010). Recent con-ceptual models have argued that experiences of

SES, Cognitive Stimulation, and Executive Function 3

threat (e.g., violence) and experiences of deprivation(e.g., an absence of cognitive stimulation) may havedistinct impacts on cognitive and neural develop-ment and that controlling for co-occurring expo-sures is critical for isolating the effects of distincttypes of environmental experience (McLaughlin,Sheridan, & Lambert, 2014; Sheridan & McLaugh-lin, 2014). To ensure that our findings reflect SESdifferences that are not explained by exposure toother forms of adversity, all analyses controlled forchildren’s exposure to violence.

Method

Participants

A sample of 101 youths aged 60–75 months(Mage = 5.55 � 0.37, 51 females) and their parentswere participated in the study between February2016 and September 2017. Families were recruitedfrom the Seattle area via fliers posted at preschools,day cares, clinics, and from the general community.To ensure SES-related diversity, recruitment effortsfocused on neighborhoods with wide variability inSES composition. The race and ethnicity of the fam-ilies closely matched the demographics of thegreater Seattle area (67.3% White, 14.8% Black, 2.9%American Indian or Alaska Native, 12.8% Asian,0.9% Native Hawaiian or Pacific Islander, 0.9%Other; 8.9% Hispanic or Latino). The InstitutionalReview Board at the University of Washingtonapproved all procedures. Participants were compen-sated and written informed consent was obtainedfrom legal guardians. Youths provided verbalassent. Two female participants were excluded fromall analyses due to having scores of verbal intelli-gence as assessed by the Peabody Picture Vocabu-lary Test (Dunn & Dunn, 2007) two standarddeviations below the mean, which was an exclusioncriteria for participation.

Socioeconomic Status

SES was assessed using two measures: theincome-to-needs ratio and maximum parental edu-cation. The income-to-needs ratio captures theamount of annual income that a family earns rela-tive to the federal poverty line for a family of thatsize. Parents reported annual income in 10 bins,and the median of the income bins was usedexcept for the lowest and highest bins, whichwere assigned $5,000 and $200,000, respectively.Income-to-needs ratio was calculated by dividingthe total household income by the 2016 U.S.

census-defined poverty line for a family of thatsize, with a value less than one indicating incomebelow the poverty line. Median income-to-needswas 4.49 with 8% of participants (income toneeds < 1) were living in poverty and 23% of par-ticipants living at less than twice the poverty line.Income-to-needs is based on the federal povertyline and does not account for regional variation incost of living. In the greater Seattle area, wheredata were collected, a 2017 study found that afamily of four requires an income of approxi-mately $75,000 per year in order afford basicneeds (i.e., food, housing, transportation, healthcare, and child care; Pearce, 2017). A family ofthree requires approximately $70,000 per year anda family of two requires approximately $57,000per year. According to this standard, nearly halfof our sample (48.5%) is below or near the self-sufficiency standard for the geographic regiontested. Income-to-needs values were log-trans-formed for all analyses, which is common indevelopmental studies (Noble et al., 2015; Rosenet al., 2018) and reflects the hypothesis that SESassociations with cognitive development are stron-gest at the lower end of the SES spectrum.

We additionally used caregiver education asanother measure of SES, coded as total years ofeducation obtained by the caregiver with the great-est educational attainment (10–22 years).

Procedure

Assessments at Time 1 (T1) took place in the partic-ipant’s home where children completed the battery ofEF tasks. Observations of the home environment werealso conducted and caregivers provided demographicinformation, including SES, and information on vio-lence exposure. A longitudinal follow-up (T2) wascompleted an average of 18 months after the T1assessment (M = 17.45 months, SD = 4.03), 76 partic-ipants (75.2% of the baseline sample) performed thesame EF tasks again in the laboratory.

HOME Assessment

Two experimenters visited the family home inorder to assess enrichment of the home environ-ment using the Home Observation of the Environ-ment (HOME), Early Childhood version (Bradleyet al., 2001). The HOME is made up of both obser-vations by the experimenter and interview ques-tions directed at the parent and a point is given forevery item coded as present. The observationcomponent includes information about what the

4 Rosen et al.

interviewer sees in the home (e.g., books, toys),observations about the parent (e.g., parent’s lan-guage use), and observations about parent–childinteractions (e.g., whether the parent responds ver-bally to the child’s questions). The interview por-tion contains questions about items the child mighthave (e.g., puzzles), questions about parent behav-iors (e.g., parent encourages child to learn numbers)and questions about parent–child interactions (e.g.,parent encourages child to talk and takes time tolisten).

We extracted one subscale from the HOME itemsfor further analysis: cognitive stimulation. Severalof the original subscales in the HOME assessment(Language Stimulation, Academic Stimulation, Vari-ety, and Learning Materials) include items reflectingcognitive stimulation. Moreover, some of these sub-scales include items that reflect other aspects of thehome environment that reflect constructs other thancognitive stimulation (e.g., parent’s voice conveyspositive feelings about child, which reflectswarmth). As such, we performed a confirmatoryfactor analysis of the HOME items based on themodel of the types of experiences underlying cogni-tive stimulation—including environmental complex-ity, enriching experiences, interactions withcaregivers, and linguistic experience (Rosen et al.,2019). Cognitive stimulation was made up of 20items that assessed learning materials and complexstimuli for the child in the home (e.g., the numberof books in the home, access to toys that teachnumbers), the variety of experiences (e.g., beingtaken to a museum in the last year, being taken ona trip at least 50 miles away within the last year),language in the home (e.g., whether parent usescomplex sentence structure or grammar) and care-giver involvement in the child’s learning (e.g., childis encouraged to learn to read a few words, child isencouraged to learn colors). Confirmatory factoranalysis indicated that our model of the constructsrepresented in the HOME items fit the data well(RMSEA < .001, 95% CI [0.000, 0.037]; Tucker–Lewis index = 1.00; comparative fit index: 1.022).See Supporting Information for information on thespecific items were included in the cognitive stimu-lation subscale. Cognitive stimulation was alsoassessed at Time 2 using a modified version of theHOME short form (Mott, 2004; Rosen et al., 2018).

Language

Although we conceptualize language exposure asa critical element of cognitive stimulation, linguisticexperience is measured in a relatively cursory

manner in the HOME assessment. Thus, we used anadditional task to assess linguistic quantity and qual-ity. Partway through the session, the caregiver andchild took a 10-min snack time break that was videorecorded. The caregiver was instructed to have aconversation with the child in the same way thatthey normally would during a snack or meal. Con-versations were then transcribed and processed withSystematic Analysis of Language Transcripts soft-ware (Salt Software LLC, Madison, WI, USA). Toassess language quantity, we used the total numberof parent words used during the interaction. Toassess language quality, we measured the total num-ber of different words, which is an assessment of thediversity of language to which the child is exposed,and the mean length of utterance of the parent,which is a measure of language complexity (Daneriet al., 2018; Hughes & Ensor, 2008). These measureshave been used in other studies to assess languagequantity and quality in young children (Daneri et al.,2018; Rowe, 2012). One conversation was unintelligi-ble due to excessive background noise and could notbe transcribed; this subject was excluded from analy-ses including language.

Behavioral Measures

Working Memory

Working memory was assessed using a child-friendly version of the backwards digit span task,which has been standardized for children in thisage range (Carlson & Meltzoff, 2008). Childrenwere told they would be playing a game wherethey say things backwards. They were then intro-duced to an Ernie doll (Sesame Street), for whomthe experimenter used a different voice. The experi-menter then did an example round with Erniewhere the experimenter said two numbers out loud,and Ernie said the string of two numbers presentedby the experimenter backwards. Participants thenunderwent practice trials with two numbers. Oncethe participants successfully completed one practiceround, they moved onto the test trials. If partici-pants did not successfully complete a practiceround, they were given scripted feedback and addi-tional instructions on how to complete the task. Ifparticipants did not successfully complete a practiceround after four trials, they did not move onto thetest trials and received a score of zero. The test tri-als consisted of four levels of increasing difficulty(two-digit, three-digit, four-digit, five-digit) of threetrials each. The experimenter presented each trial ina steady tone of voice and the participant’s

SES, Cognitive Stimulation, and Executive Function 5

response was recorded. If the participant completedat least one correct trial, they proceeded to the nextlevel. Participants received a point for each correcttrial.

Inhibition

To assess inhibition we used a standard test ofSimon Says in which participants were instructedto imitate the experimenter’s action if the actionwas proceeded with “Simon says” and to inhibittheir response when this phrase was not uttered(Carlson & Meltzoff, 2008). After the rules of thegame were introduced, participants responded to aseries of questions about the rules to ensure theycomprehended them. Participants underwent 10imitation and 10 inhibition trials, which were inter-mixed. For imitation trials, participants receivedthree points for each successfully completed action,two points for each partial action, one point for aflinch or wrong movement, and zero point for nomovement. For inhibition trials, participantsreceived three points for no movement, two pointsfor a flinch or wrong movement, one point for apartial movement, and zero point for a completemovement. Participants who were unable to passthe practice after four rounds of reminders of therules were given a score of zero for both imitationand inhibition. Performance was scored by tworaters from video recordings and discrepancieswere resolved among the two raters; inter-rater reli-ability was good (Cohen’s kappa T1: .76, Cohen’skappa T2: .84).

Cognitive Flexibility

To assess cognitive flexibility, we administereda child-friendly version of a Dimensional CardSorting Task that has been standardized for chil-dren in this age range (Zelazo, 2006) in which chil-dren were instructed to sort cards based on coloror shape. A box with a blue star and a box with ared truck were placed in front of the participant.Subjects were presented with cards with bluetrucks and red stars. During the first round (pre-switch), subjects were instructed to sort the cardsinto the appropriate box based on the color of theshape on the card (five color trials). During thesecond round (post-switch), the rule switched andparticipants were instructed to sort the cards byshape (five shape trials). In the third round(switching), the experimenter verbally instructedthe participant to sort by shape or by color beforeeach trial (five color trials, five shape trials). In the

final round, participants were presented with somecards that had a colored border and some thathad no border. Participants were instructed thatthey should sort by color if the card had a border,and sort by shape if they had a card with no bor-der (five color trials, five shape trials). Subjectsmoved on to the next round if they got one orfewer wrong answers on the color or shape trialsfor each level. Participants were then given onepoint for each level passed for a maximum of fourpoints. One male subject elected not to perform EFtasks and was excluded from analysis includingEF measures.

Academic Achievement

During the T2 follow-up, three subsets of theWoodcock–Johnson IV Tests of Achievement wereused as assessments of academic achievement(Schrank, Mather, & McGrew, 2014): Letter-WordIdentification, Spelling, Calculation. Each test pre-sented the participants with items of increasingdifficulty. In the Letter-Word Identification test,participants were asked to identify letters andread lists of words. In the Spelling test, partici-pants were instructed to spell words that wereread aloud and used in a sentence by the experi-menter. The Calculation test required children tocomplete a series of arithmetic problems. The Let-ter-Word Identification, Spelling, and Calculationsubsets were all discontinued when the partici-pants answered incorrectly on six consecutiveitems. Standard scores normed by age were calcu-lated for each subset as measures of the child’sachievement in that academic domain and theAcademic Skills Cluster was calculated based onthese scores.

Violence Exposure

To assess exposure to violence, parents com-pleted the Violence Exposure Scale for Children–Revised (VEX–R; Fox & Leavitt, 1995) in a formatadapted for parent rather than child report. Thisassessment measures the frequency that a child haswitnessed violence (e.g., seeing someone be hitreally hard; witnessing someone be stabbed or shot)and directly experienced violence (e.g., being beatenup, being pushed or shoved). A total score reflect-ing the frequency of experiencing or witnessing vio-lence was created by summing the items, and thisvariable was included as a covariate in all analyses.All analyses presented in the manuscript used theVEX–R as a covariate.

6 Rosen et al.

Statistical Analyses

Statistical analyses were performed in SPSS 20(IBM Corp, Armonk, NY, USA). We had two over-arching goals. The first was to examine the role ofcognitive stimulation as a mechanism linking child-hood SES with EF. To do so, we first tested each ofthe paths of a standard mediation model. First, weused linear regression to examine the association ofSES and EF performance at T1 and at T2, control-ling for T1 performance. Specifically, we estimateda series of separate multivariate models examiningincome-to-needs and parental education as predic-tors of working memory, inhibition, and cognitiveflexibility performance (c path). Next we examinedthe associations of the two SES measures with cog-nitive stimulation based on the HOME assessment(i.e., our cognitive stimulation factor) and languageexposure (i.e., language quantity using total num-ber of words, and mean length utterance in words,and language quality using number of differentwords) during the snack time conversation (a path).Finally, we examined the associations of our mea-sures of cognitive stimulation with performance oneach of the EF tasks at T1 and T2, controlling forT1 performance (b path). All analyses controlled forage, sex, and violence exposure. Residualizedchange in EF from T1 to T2 was estimated in alllongitudinal models by controlling for performanceat T1.

The second goal was to examine whether EF atT1 explained SES-related differences in academicachievement at T2. As such, we additionally testedthe associations between SES (income and parentaleducation) and academic achievement as well asthe associations between all three measures of EF atT1 with academic achievement at T2.

All results were false discovery rate (FDR) cor-rected at the level of hypothesis (e.g., to test thehypothesis that SES is related to EF, we performedsix tests, so we FDR corrected for those six tests, toa corrected p-value of .05).

Mediation

After testing each of these paths, we used a stan-dard test of statistical mediation that estimates thesignificance of indirect effects using a bootstrappingapproach that provides confidence intervals for theindirect effects (Hayes, 2013). Confidence intervalsthat do not include 0 are considered evidence forstatistically significant indirect effects. We tested theindirect effect for factors significantly associatedwith both SES and EF.

Sensitivity Analyses

We additionally performed sensitivity analysescontrolling for child verbal intelligence as measuredby the Peabody Picture Vocabulary Test (PPVT) todetermine whether the associations between SESand EF and cognitive stimulation and EF persistedafter accounting for verbal ability.

Results

Descriptive Statistics

Means and standard deviations for all studyvariables are presented in Table 1, and bivariatecorrelations between all study variables are pre-sented in Table 2.

SES and EF (c Path)

First, we assessed the association between SESand EF at T1. Family SES was positively associatedwith performance on all three EF tasks. Specifically,

Table 1Means and Standard Deviations of All Study Variables

Measure M (SD) Range

Income $112,530 $64,961 $5,000–$250,000Income-to-needs 4.73 2.86 .08–10.5Education 4.04 1.05 10–22Violence exposure 3.00 3.90 0–20Cognitive stimulation

(total score)15.69 3.07 5–20

Total number of words 662.07 221.81 291–1,270Mean length utterance 4.57 0.81 1.71–7.10Total number of different

words212.44 48.83 43–331

Backwards digit span totalpoints (Time 1)

3.68 2.04 0–8

Backwards digit span totalpoints (Time 2)

5.68 1.80 2–11

Simon says inhibition totalpoints (Time 1)

17.31 10.02 0–30

Simon says inhibition totalpoints (Time 2)

24.04 5.39 0–30

Dimensional card sorthighest level passed(Time 1)

2.91 0.90 1–4

Dimensional card sorthighest level passed(Time 2)

3.57 0.68 1–4

Academic achievement 100.42 13.39 71–141

SES, Cognitive Stimulation, and Executive Function 7

Tab

le2

Correlatio

nsof

AllStud

yVariables

Age

(T1)

Age

(T2)

Sex

Violenc

eItN

(Log

)Edu

CS

TW

MLU

NDW

BDS

(T1)

BDS

(T2)

Simon

(T1)

Simon

(T2)

DCCS

(T1)

AA

Age

(T1)

Age

(T2)

.702**

Sex

.007

�.00

1Violenc

e.063

�.03

1.010

ItN

(Log

).078

<.001

�.01

6�.

345**

Edu

�.01

4.073

�.09

3�.

359**

.493**

CS

�.03

0.014

.070

�.17

4.474**

.528**

TW

�.00

5�.

107

.138

.130

�.07

2.094

�.06

0MLU

.064

.067

�.01

0�.

017

.067

.304*

.194

.611**

NDW

.045

�.06

0.179

.115

.010

.164

.035

.932**

.640**

BDS(T1)

.214*

.177

.082

�.39

1**

.418**

.337**

.447**

�.06

0.150

�.00

8BDS(T2)

.391**

.323**

.085

�.27

6*.356**

.364**

.306**

.040

.201

�.12

6.523**

Simon (T1)

.035

.162

.013

�.31

7**

.333**

.331**

.380**

.078

.245*

.147

.353**

.323**

Simon (T2)

�.07

3.152

.216

�.07

1.057

.118

.337**

�.12

5.103

�.07

6.133

.320**

.334**

DCCS(T1)

.108

.125

�.00

2�.

279**

.339**

.291**

.420**

�.09

8.185

�.03

3.510**

.351**

.458**

.232*

DCCS(T2)

.063

.239*

�.02

8�.

099

.214

.238*

.386**

�.22

4.058

�.13

9.266*

.266*

.254*

.134

.481**

AA

�.02

2�.

167

�.12

6�.

264*

.279*

.267*

.121

�.13

0.006

�.06

5.390**

.324**

.088

�.00

3.312**

.177

Note.

T1=Tim

e1;

T2=Tim

e2;

Violenc

e=totalscorereflectin

gthefreq

uenc

yof

expe

rien

cing

violen

ceas

assessed

bytheViolenc

eExp

osureScaleforChildren–

Rev

ised

(VEX

–R);

ItN

=income-to-needsratio

;Edu=high

estleve

lof

parental

educ

ation;

CS=cogn

itive

Stim

ulationas

assessed

bytheHom

eObserva

tionof

theEnv

iron

men

t;TW

=totalwords

;MLU

=meanleng

thutteranc

ein

words;NDW

=nu

mbe

rof

differen

twords;BDS=ba

ckwardsdigitsp

anscoreto

assess

working

mem

ory;

Simon

=Simon

Says

scoreon

inhibitio

ntrials

toassess

inhibitio

n;DCCS=Dim

ension

alCha

ngeCardSo

rthigh

estleve

lpa

ssed

toassess

cogn

itive

flexibility;AA

=acad

emic

achiev

emen

tas

measu

redby

theW

oodc

ock–

John

sonAcadem

icSk

illsCluster.

*p<.05.

**p<.01.

8 Rosen et al.

income-to-needs and parent educational attainmentwere associated with working memory performanceon the backwards digit span task (b = .299,p = .012; b = .234, p = .019, respectively, Figures 1Aand 1B), inhibition as measured by the Simon Saystask (b = .251, p = .019; b = .252, p = .019, respec-tively, Figures 1C and 1D), and cognitive flexibilityas measured by the dimensional card sorting task(b = .264, p = .019; b = .219, p = .037, respectively,Figures 1E and 1F). Next, we examined the associa-tions between SES and growth in EF over time.After correction for multiple comparisons, neithermeasure of SES was associated with change in EFperformance from T1 to T2 (ps > .18), although par-ental education was significantly associated withgrowth in working memory before FDR correction(b = .219, p = .030, uncorrected).

SES and Cognitive Stimulation (a Path)

Next, we assessed the association between SESand cognitive stimulation at T1. There was a strongpositive association between both income-to-needsand parental education with cognitive stimulationas measured by the HOME assessment of (b = .478,p < .001; b = .547, p < .001, respectively, Figures 2Aand 2B).

With regard to linguistic experience, we foundsome evidence for differences in quality, but notquantity, of language exposure as a function ofeducation but not income-to-needs. Specifically, wefound that parental education predicted the meanlength utterance (b = .294, p = .006), whereasincome-to-needs did not (b = .068, p = .670). Therewas a trend toward an association between

Figure 1. Linear regression between socioeconomic status and working memory (A and B), inhibition (C and D), and cognitive flexibil-ity (E and F), controlling for age, sex, and violence exposure at T1. All p-values are FDR corrected.

SES, Cognitive Stimulation, and Executive Function 9

education and number of different words (b = .217,p = .094,), but no significant association betweenincome and number of different words (b = .047,p = .670). Neither measure of SES was associatedwith language quantity as measured by total num-ber of words (b = �.012, p = .912, b = .140, p = .412for income-to-needs and education, respectively).

Cognitive Stimulation and EF (b Path)

Cognitive stimulation as measured by the HOMEassessment was positively associated with all threemeasures of EF at T1. Specifically, greater cognitivestimulation in the home was positively associatedwith working memory performance on backwardsdigit span (b = .392, p < .001, Figure 3A), inhibitionduring Simon Says (b = .337, p = .001, Figure 3B),and cognitive flexibility on the dimensional cardsorting task (b = .388, p < .001, Figure 3C). Violenceexposure was negatively associated with perfor-mance in these models (b = �.338, p = .001 forworking memory, b = �.263, p = .009 for inhibition,and b = �.220, p = .019), although these associa-tions were consistently smaller than for cognitivestimulation (see Table S1).

Next, we tested the association between linguis-tic experience and EF at T1. Language complexityas measured by mean length utterance was margin-ally associated with inhibition after FDR correction(b = .242, p = .078), but not associated with work-ing memory (b = .133, p = .225) or cognitive flexi-bility (b = .177, p = .148). Language variety asmeasured by number of different words was notassociated with EF (b = .013, p = .938; b = .186,p = .148; b = �.008, p = .938, for working memory,inhibition, and flexibility, respectively). We also

investigated whether language quantity as mea-sured by total number of words was associatedwith EF and found no significant associations withworking memory, inhibition, or cognitive flexibility(b = �.18, p = .852; b = .122, p = .675; b = �.063,p = .801, respectively).

We then tested whether cognitive stimulationwas associated with growth in EF over time (T2performance controlling for T1 performance). AfterFDR correction, cognitive stimulation, as measuredby the HOME assessment was associated withgrowth in cognitive flexibility (b = .268, p = .054)and marginally associated with growth in inhibition(b = .224, p = .087), but not with growth in workingmemory (b = .141, p = .18). Violence exposure wasnot associated with growth in EF in any of thesemodels (ps > .42). Neither language quantity noreither measure of language quality was associatedwith growth in any measure of EF (ps > .28), norwas language quantity associated with growth inany measure of EF (ps > .12). Additionally, EF atT1 did not predict changes in cognitive stimulationmeasured at T2, controlling for T1 cognitive stimu-lation (ps > .5), which is inconsistent with the ideathat higher EF is driving higher cognitive stimula-tion from parents.

Mediation Analyses (c0 Path)

Finally, we conducted mediation analyses todetermine whether the degree of cognitive stimula-tion in the home environment mediated the associa-tion between family SES and EF (Figure 4).Consistent with our hypotheses, we found a signifi-cant indirect effect of income-to-needs (95% CI[0.15, 0.62]) and parental education (95% CI [0.09,

Figure 2. Linear regression between two measures of SES (A) Log Income-to-Needs and (B) Education, with cognitive stimulation inthe home environment controlling for age, sex, and violence exposure at T1.

10 Rosen et al.

0.24]) on working memory performance andcognitive flexibility (95% CI [0.07, 0.27] for income-to-needs and 95% CI [0.03, 0.11] for parental educa-tion) through cognitive stimulation. We also founda significant indirect effect of income-to-needs (95%CI [0.42, 2.90]) on inhibition through cognitive stim-ulation. Furthermore, there was a significant indi-rect effect of parental education on inhibitionthrough cognitive stimulation and mean lengthutterance (95% CI [0.07, 1.09]).

Sensitivity Analyses

We conducted sensitivity analyses to determinewhether these associations persisted after control-ling for verbal ability as measured by the PPVT.

The findings were largely unchanged. Briefly, cog-nitive stimulation is associated with performanceon all three tests of EF at T1 after controlling forverbal ability. Both measures of SES have indirecteffects on all three measures of EF through cogni-tive stimulation at T1. Additionally, there is a mar-ginally significant association between cognitivestimulation and growth in both inhibition and cog-nitive flexibility at T2 after controlling for verbalability. Detailed results are presented in theTable S2.

SES and Academic Achievement

We next assessed whether SES at T1 was associ-ated with academic achievement at T2. SES was

Figure 3. Linear regression between cognitive stimulation and working memory (A), inhibition (B), and cognitive flexibility (C), control-ling for age, sex, and violence exposure at T1. All p-values are FDR corrected.

Figure 4. Mediation models. Cognitive stimulation fully mediated the associations between socioeconomic status (income-to-needs andeducation) and executive functions (working memory, inhibition, and cognitive flexibility) at T1. Coefficients are unstandardized.*p < .05. **p < .01.

SES, Cognitive Stimulation, and Executive Function 11

marginally associated with academic achievement(b = .222 p = .061 and b = .215, p = .061, for incomeand education, respectively).

EF and Academic Achievement

We next assessed the associations between EF atT1 and academic achievement at T2. Working mem-ory and cognitive flexibility were both positivelyassociated with academic achievement (b = .398,p = .003 and b = .286, p = .017, respectively),whereas inhibition was not (b = .141, p = .666).

Mediation Analyses

Finally, we conducted mediation analyses todetermine whether EF explained SES-related differ-ences in academic achievement (Figure 5). Wefound that working memory significantly mediatedthe association between income-to-needs and aca-demic achievement (95% CI [0.42, 4.03]). At a moreliberal threshold, working memory mediated theassociation between education and achievement(90% CI [0.004, 0.89]). Cognitive flexibility alsomediated the association between income-to-needsand education with achievement at a more liberalthreshold (90% CI [0.11, 2.20] and 90% CI [0.01,0.64], respectively).

Discussion

This study adds to a small but growing literaturehighlighting an important role of cognitive stimula-tion in the early home environment in the develop-ment of EF. We investigated cognitive stimulation—assessed with observational measures of environ-mental complexity, caregiver interactions, and lan-guage quality and quantity—as a mechanismexplaining SES-related differences in the

development of EF in children. Consistent with pre-vious studies, SES was associated with workingmemory, inhibition, and cognitive flexibility (Dil-worth-Bart, 2012; Hackman & Farah, 2009; Lawsonet al., 2018; Noble et al., 2007; Rosen et al., 2018).At T1, SES was also strongly associated with cogni-tive stimulation in the home environment, such thatincome-to-needs and parental education were posi-tively associated with our observational measure ofcognitive stimulation; parental education was addi-tionally associated with parent language quality.Cognitive stimulation, in turn, was positively asso-ciated with working memory, inhibition, and cogni-tive flexibility, whereas language quality wasspecifically associated with inhibition. Consistentwith our hypotheses, cognitive stimulation medi-ated the concurrent associations at T1 between bothmeasures of SES and all three measures of EF.Moreover, cognitive stimulation at T1 was signifi-cantly associated with growth in inhibition and cog-nitive flexibility over an 18-month follow-upperiod, whereas SES was not associated withgrowth in EF. Critically, our results remained lar-gely unchanged when we included verbal intelli-gence as a control variable in our analyses. Thesefindings provide the first longitudinal evidenceusing observational assessment of the home envi-ronment indicating that cognitive stimulation in thehome environment is associated with the develop-ment of two core aspects of EF. The significantassociations of cognitive stimulation with growth inEF over time are notable, given that SES associa-tions with EF emerge early and remain relativelystable over time (Hackman et al., 2015; Lenguaet al., 2015). We additionally demonstrate that vari-ation in working memory and cognitive flexibilitypredicts future academic achievement, and thatworking memory and flexibility mediate the associ-ation between income and achievement. These find-ings suggest that cognitive stimulation may be animportant target for interventions aimed at reduc-ing the SES gap in EF and that the resultingimprovements in EF may have a downstreamimpact on academic achievement.

Here, we replicate and extend previous studiesdemonstrating that cognitive stimulation is a mech-anism explaining SES-related differences in EF. Sar-sour et al. (2011) found that exposure to enrichingactivities—an aspect of cognitive stimulationincluded in this study—mediated the cross-sectionalassociation between SES and working memory andinhibition in older children, aged 8–12 years. Fur-thermore, recent work from Amso et al. (2018)demonstrated that cognitive stimulation mediated

Figure 5. Academic achievement. Working memory fully medi-ated the association between income-to-needs and academicachievement. Coefficients are unstandardized. †p < .1. *p < .01.**p < .05.

12 Rosen et al.

the association between SES and working memory.We extend these cross-sectional findings by demon-strating that cognitive stimulation is associated withgrowth in EF during early childhood. The only priorlongitudinal study on this topic found that cognitivestimulation as measured by parent report of learningmaterials, variety of experiences, and academic stim-ulation mediated the association between SES andworking memory and planning (Hackman et al.,2015). We extend this prior work using observationalmeasures of cognitive stimulation and by document-ing the mediating role of cognitive stimulation in thelink between SES and two additional aspects of EF:inhibition and cognitive flexibility (Miyake et al.,2001). We further extend this work by demonstratingthat cognitive stimulation in the home environmentis associated with growth in EF over time.

Consistent with other studies we demonstrate thatcognitive stimulation mediates SES-related differ-ences in working memory performance measuredconcurrently (Amso et al., 2018; Sarsour et al., 2011).However, we did not find that cognitive stimulationpredicted growth in working memory in an 18-month follow up. Given that recent evidence sug-gests that cognitive stimulation plays an importantrole in explaining SES-related differences in workingmemory performance in older children and adoles-cents (Amso et al., 2018), one possibility is that thereare developmental differences in the importance ofcognitive stimulation across the different compo-nents of EF. However, future longitudinal studiesbeginning earlier in development would be neededto address this question.

The findings of this study highlight that the homeenvironment has a pronounced role in the develop-ment of cognitive abilities in early childhood. Cogni-tive stimulation in school and other environments islikely important for the development of EF, but earlyin development the most proximal context is in thehome environment. As children develop and spendmore time in other contexts, however, the impor-tance of cognitive stimulation outside the home mayincrease (Crosnoe et al., 2010). Therefore, future lon-gitudinal studies should examine the role of cogni-tive stimulation in the classroom as an additionalmechanism underlying growth in EF over timeamong school-aged children.

This study also extends this previous work byincluding both a measure of parental language, acritical aspect of cognitive stimulation. Our findingsare somewhat consistent with a recent study thatfound maternal language complexity mediated theassociation between SES and a composite score ofchild EF (Daneri et al., 2018). In contrast, we found

a specific link between language complexity, asmeasured by mean length utterance, and inhibition,but not working memory or cognitive flexibility.Although the mechanisms underlying this associa-tion are unknown, one possibility is that greatercomplexity of parent language may facilitate thedevelopment of inhibition in children by requiringthem to suppress a response for a longer period oftime during a conversation to wait for the speakerto complete their turn (McLaughlin, 2016). Greatercomplexity of language exposure may also increasechildren’s ability to internally represent regulatoryspeech, which could be used to help with inhibitionof a prepotent response (Peterson, Bates, & Staples,2015; Valloton & Ayoub, 2011).

Considerable evidence suggests that cognitivestimulation provides the building blocks for devel-opment of EF in childhood. Given the meaningfulrole that caregivers play in shaping the amount ofcognitive stimulation children experience early indevelopment, it has been argued that in environ-ments with limited caregiver interactions, childrenhave less external guidance to regulate attention andhave fewer experiences that require attention to besustained (Rosen et al., 2019). This reduced caregiverinteraction coupled with reduced access to sensorycomplexity (e.g., reduced access to books, toys, andcomplex stimuli with which to engage) may result indelayed development of attention regulation mecha-nisms and language ability in children. The develop-ment of these lower order cognitive functions may inturn, scaffold development of higher order cognitiveabilities including EF. A recent study highlightedother lower order functions, such as children’s expe-riences in directing their attention in anticipation ofimpending events, as building blocks for EF (Weiss,Meltzoff, & Marshall, 2018).

Indeed, it is well-established that far moreextreme environments lacking in cognitive stimula-tion and caregiver interaction, such as institutionalrearing and neglect, are associated with large andlasting difficulties with EF (Bos, Fox, Zeanah, &Nelson, 2009; Loman et al., 2013; McLaughlin,Sheridan, & Nelson, 2017; Tibu et al., 2016). Evenchildren who are removed from these types ofdeprived environments and placed in a more cogni-tively stimulating environment before the age of24 months exhibit attentional impairments in latechildhood and early adolescence (Slopen, McLaugh-lin, Fox, Zeanah, & Nelson, 2012; Tibu et al., 2016).Furthermore, recent evidence suggests that increas-ing cognitive stimulation in the home is an effectivestrategy in improving EF in childhood. A random-ized controlled trial in Pakistan found that an

SES, Cognitive Stimulation, and Executive Function 13

intervention designed to improve cognitive stimula-tion in the home environment was associated withgains in EF skills over time and this effect wasstronger than a nutrition intervention (Yousafzaiet al., 2016). Together, this work highlights the criti-cal role of cognitive stimulation in the home in sup-porting the development of EF.

It is notable that our measure of cognitive stimu-lation included both items that reflect access toresources that are important for learning (e.g., Childhas toys that teach colors) as well as items thatreflect parental engagement in learning (e.g., Childis encouraged to learn numbers). We believe thataccess to learning materials coupled with parentalengagement in learning together scaffold devel-opment of EF in children. It is unlikely that SES-related differences in EF would be mitigated bysimply providing families with learning materialswithout simultaneously providing parents withinformation on effective ways of engaging withtheir child as well as work to eliminate the struc-tural barriers that constrain time for parent-childinteractions. However, future studies would beneeded to test this hypothesis directly.

Recent work has sought to disentangle how dis-tinct dimension of childhood adversity, includingdeprivation and threat, may be associated with dif-ferent cognitive and neural outcomes (McLaughlinet al., 2014; Sheridan & McLaughlin, 2014). Thisstudy sought to examine the associations of cogni-tive stimulation with EF after controlling for expo-sure to violence. We hypothesized that cognitivestimulation would be the environmental factormost strongly related to EF and growth in EF overtime, and our results support this hypothesis. Wealso found that violence exposure was associatedwith EF at baseline, though to a lesser degree thancognitive stimulation. A cross-sectional associationof violence with EF is consistent with some previ-ous work (Hanson et al., 2010) but contrasts withseveral recent studies in large sample demonstrat-ing an absence of association between violenceexposure and EF—including working memory,inhibition, and cognitive flexibility—in adolescentsafter controlling for SES (e.g., Lambert et al., 2017;Sheridan, Peverill, Finn, & McLaughlin, 2017). Thefact that we observed residual associations of vio-lence exposure with EF in our sample of youngchildren may suggest that violence exposure has amore powerful effect on EF early in developmentor that these effects become weaker across develop-ment (but see Machlin, Miller, Snyder, McLaughlin,& Sheridan, 2019). Future studies with samplesspanning a wider age range are needed to evaluate

this possibility empirically. Critically, only cogni-tive stimulation was associated with growth in EF,suggesting that interventions targeting cognitivestimulation are likely to be most effective in miti-gating SES related differences in the developmentof EF.

This study has several limitations that should beacknowledged. First, our language assessment wasshort in length (10 min) and potentially limited inits content. Parents often took a few minutes to getcomfortable and conversations often focused ontalking about the prizes children had just won orasking about the games they had played. Therefore,it is likely that these conversations did not fullycapture a natural everyday snapshot of languageexposure. A more open-ended conversational per-iod like the book-sharing task employed by Daneriet al. (2018) might provide a more realistic pictureof language exposure in the home. Alternatively,Language Environment Analysis (LENA) technol-ogy can track language exposure over a 16 hr per-iod and provide a potentially more representativesample of language exposure and language expo-sure has been shown to vary by SES using LENA(Gilkerson et al., 2017; Ram�ırez-Esparza et al., 2014;Romeo et al., 2018). Indeed, others have recentlyfound that children’s language exposure and expe-rience as measured by LENA is predictive of lan-guage ability and prefrontal cortex function (Romeoet al., 2018). Future studies should employ thesetools to explore the role of language exposure inSES-related differences in working memory, inhibi-tion, and cognitive flexibility. Second, we had a rel-atively large range in time between the first waveand the follow-up time. This range limits the preci-sion of this study and future studies should workto have more precise timing between study waves.Third, while our sample was diverse with respectto income, the sample was relatively highly edu-cated. While we still found significant associationsbetween education and both cognitive stimulationand EF and results were consistent across measuresof SES, future studies should work to replicate thepresent findings with a more educationally diversesample. Fourth, EF develops rapidly during thisperiod of development and as such, another limita-tion is that that ceiling effects may have constrainedvariability in inhibition and cognitive flexibilityscores at T2. Finally, while the present findingsextend previous work that cognitive stimulation inthe home explains SES related differences in EF inyounger children, many studies have demonstratedSES-related differences in EF emerge much earlierin development (e.g., Lengua et al., 2015).

14 Rosen et al.

Therefore, future studies should aim to replicatethese findings in an even younger sample.

Conclusions

This study highlights the important role that cog-nitive stimulation plays in the development of EFand that differences in working memory play ameaningful role in explaining SES-related differ-ences in academic achievement in early childhood.These findings along with other recent studies(Daneri et al., 2018; Hackman et al., 2015; Sarsouret al., 2011; Yousafzai et al., 2016) point to cognitivestimulation as a plausible and modifiable environ-mental mechanism that contributes to these SES-related differences in cognitive development.Additionally, understanding how cognitive stimula-tion impacts the brain systems that support EF mayshed light onto the neural mechanisms underlyingSES-related differences in cognitive development.Indeed, recent work has shown that cognitive stim-ulation is associated with greater cortical thicknessin the frontoparietal network (Rosen et al., 2018).Together with this study, these findings point tocognitive stimulation as an important mechanismthat contributes to SES-related disparities in cogni-tive development, particularly EF. Interventionsthat target cognitive stimulation may be promisingfor reducing SES-related disparities in both EF andacademic achievement.

References

Amso, D., Salhi, C., & Badre, D. (2018). The relationshipbetween cognitive enrichment and cognitive control: Asystematic investigation of environmental influences ondevelopment through socioeconomic status. Develop-mental Psychobiology, 61, 159–178. https://doi.org/10.1002/dev.21794

Blair, C. (2002). School readiness. American Psychologist,57, 111–127. https://doi.org/10.1037//0003-066X.57.2.111

Blair, C., & Razza, R. P. (2007). Relating effortful control,executive function, and false belief understanding toemerging math and literacy ability in kindergarten.Child Development, 78, 647–663. https://doi.org/10.1111/j.1467-8624.2007.01019.x

Bos, K. J., Fox, N., Zeanah, C. H., & Nelson, C. A. (2009).Effects of early psychosocial deprivation on the devel-opment of memory and executive function. Frontiers inBehavioral Neuroscience, 3, 16. https://doi.org/10.3389/neuro.08.016.2009

Bradley, R. H., & Corwyn, R. F. (2002). Socioeconomicstatus and child development. Annual Review of

Psychology, 53, 371–399. https://doi.org/10.1146/annurev.psych.53.100901.135233

Bradley, R. H., Corwyn, R. F., McAdoo, H. P., & GarciaColl, C. (2001). The home environments of children inthe United States part I: Variations by age, ethnicity,and poverty status. Child Development, 72, 1844–1867.https://doi.org/10.1111/1467-8624.t01-1-00382

Carlson, S. M. (2009). Social origins of executive functiondevelopment. New Directions for Child and AdolescentDevelopment, 123, 87–98. https://doi.org/10.1002/cd.237

Carlson, S. M., & Meltzoff, A. (2008). Bilingual experienceand executive functioning in young children. Develop-mental Science, 11, 282–298. https://doi.org/10.1111/j.1467-7687.2008.00675.x

Christensen, D. L., Schieve, L. A., Devine, O., & Drews-Botsch, C. (2014). Socioeconomic status, child enrich-ment factors, and cognitive performance among pre-school-age children: Results from the follow-up ofgrowth and development experiences study. Research inDevelopmental Disabilities, 35, 1789–801. https://doi.org/10.1016/j.ridd.2014.02.003

Clearfield, M. W., & Niman, L. C. (2012). SES affectsinfant cognitive flexibility. Infant Behavior and Develop-ment, 35, 29–35. https://doi.org/10.1016/j.infbeh.2011.09.007

Coldren, J. T. (2013). Cognitive control predicts academicachievement in kindergarten children. Mind, Brain, andEducation, 7, 40–48. https://doi.org/10.1111/mbe.12006

Crosnoe, R., Leventhal, T., Wirth, R. J., Pierce, K. M.,Pianta, R. C., & NICHD Early Child Care ResearchNetwork. (2010). Family socioeconomic status and con-sistent environmental stimulation in early childhood.Child Development, 81, 972–87. https://doi.org/10.1111/j.1467-8624.2010.01446.x

Daneri, M. P., Blair, C., & Kuhn, L. J. (2018). Maternallanguage and child vocabulary mediate relationsbetween socioeconomic status and executive functionduring early childhood. Child Development. https://doi.org/10.1111/cdev.13065

Dilworth-Bart, J. E. (2012). Does executive function medi-ate SES and home quality associations with academicreadiness? Early Childhood Research Quarterly, 27, 416–425. https://doi.org/10.1016/j.ecresq.2012.02.002

Dunn, L. M., & Dunn, D. M. (2007). Peabody Picture Vocab-ulary Test 4. Summary (Vol. 30). https://doi.org/10.1037/t15144-000

Engel, P. M. J., Santos, F. H., & Gathercole, S. E. (2008).Are working memory measures free of socioeconomicinfluence? Journal of Speech Language and HearingResearch, 51, 1580. https://doi.org/10.1044/1092-4388(2008/07-0210)

Farah, M. J., Shera, D. M., Savage, J. H., Betancourt, L.,Giannetta, J. M., Brodsky, N. L., . . . Hurt, H. (2006).Childhood poverty: Specific associations with neu-rocognitive development. Brain Research, 1110, 166–174.https://doi.org/10.1016/j.brainres.2006.06.072

SES, Cognitive Stimulation, and Executive Function 15

Fernald, A., Marchman, V. A., & Weisleder, A. (2013).SES differences in language processing skill and vocab-ulary are evident at 18 months. Developmental Science,16, 234–248. https://doi.org/10.1111/desc.12019

Finn, A. S., Minas, J. E., Leonard, J. A., Mackey, A. P.,Salvatore, J., Goetz, C., . . . Gabrieli, J. D. E. (2016).Functional brain organization of working memory inadolescents varies in relation to family income and aca-demic achievement. Developmental Science, 20(5), 1–15.https://doi.org/10.1111/desc.12450

Fox, N. A., & Leavitt, N. A. (1995). The violence exposurescale for children-VEX. College Park, MA: Department ofHealth Development, University of Maryland.

Gilkerson, J., Richards, J. A., Warren, S. F., Montgomery,J. K., Greenwood, C. R., Oller, D. K., . . . Paul, T. D.(2017). Mapping the early language environment usingall-day recordings and automated analysis. AmericanJournal of Speech-Language Pathology, 26, 248–265.https://doi.org/10.1044/2016_AJSLP-15-0169

Hackman, D., & Farah, M. (2009). Socioeconomic statusand the developing brain. Trends in Cognitive Sciences,13(2), 65–73. https://doi.org/10.1016/j.tics.2008.11.003

Hackman, D. A., Farah, M. J., & Meaney, M. J. (2010).Socioeconomic status and the brain: Mechanisticinsights from human and animal research. NatureReviews Neuroscience, 11, 651–659. https://doi.org/10.1038/nrn2897

Hackman, D. A., Gallop, R., Evans, G. W., & Farah, M. J.(2015). Socioeconomic status and executive function:Developmental trajectories and mediation. Developmen-tal Science, 18, 686–702. https://doi.org/10.1111/desc.12246

Hanson, J. L., Chung, M. K., Avants, B. B., Shirtcliff, E.A., Gee, J. C., Davidson, R. J., & Pollak, S. D. (2010).Early stress is associated with alterations in the orbito-frontal cortex: A tensor-based morphometry investiga-tion of brain structure and behavioral risk. Journal ofNeuroscience, 30, 7466–7472. https://doi.org/10.1523/JNEUROSCI.0859-10.2010

Hart, B., & Risley, T. R. (1995). Meaningful differences inthe everyday experiences of young American children. Balti-more, MD: Paul H Brookes. https://doi.org/10.1007/s00431-005-0010-2

Hayes, A. (2013). Introduction to mediation, moderation, andconditional process analysis. New York, NY: Guilford.

Hughes, C. H., & Ensor, R. A. (2008). How do families helpor hinder emergence of early EF. Core Competencies toPrevent Problem Behaviors and Promote Positive YouthDevelopment, 122, 61–74. https://doi.org/10.1002/cd.234

Johnson, S. B., Riis, J. L., & Noble, K. G. (2016). State ofthe art review: Poverty and the developing brain. Pedi-atrics, 137, 1–17. https://doi.org/10.1542/peds.2015-3075

Lawson, G. M., Hook, C. J., & Farah, M. J. (2018). Ameta-analysis of the relationship between socioeco-nomic status and executive function performanceamong children. Developmental Science, 21, e12529.https://doi.org/10.1111/desc.12529

Lambert, H. K., Sheridan, M. A., Sambrook, K. A., Rosen,M. L., Askren, M. K., & McLaughlin, K. A. (2017). Hip-pocampal contribution to context encoding acrossdevelopment following early-life adversity. The Journalof Neuroscience, 37(7), 1925–1934. https://doi.org/10.1523/JNEUROSCI.2618-16.2017

Lengua, L. J., Honorado, E., & Bush, N. R. (2007). Contex-tual risk and parenting as predictors of effortful controland social competence in preschool children. Journal ofApplied Developmental Psychology, 28, 40–55. https://doi.org/10.1016/j.appdev.2006.10.001

Lengua, L. J., Kiff, C., Moran, L., Zalewski, M., Thomp-son, S., Cortes, R., & Ruberry, E. (2014). Parentingmediates the effects of income and cumulative risk onthe development of effortful control. Social Development,23, 631–649. https://doi.org/10.1111/sode.12071

Lengua, L. J., Moran, L., Zalewski, M., Ruberry, E., Kiff,C., & Thompson, S. (2015). Relations of growth ineffortful control to family income, cumulative risk, andadjustment in pre-school age children. Journal of Abnor-mal Child Psychology, 43, 705–720. https://doi.org/10.1007/s10802-014-9941-2

Lipina, S. J., Martelli, M. I., Vuelta, B., & Colombo, J.A. (2005). Performance on the A-not-B task of Argen-tinean infants from unsatisfied and satisfied basicneeds homes. Interamerican Journal of Psychology, 39,49–60.

Loman, M. M., Johnson, A. E., Westerlund, A., Pollak, S.D., Nelson, C. A., & Gunnar, M. R. (2013). The effect ofearly deprivation on executive attention in middlechildhood. Journal of Child Psychology and Psychiatry andAllied Disciplines, 54, 37–45. https://doi.org/10.1111/j.1469-7610.2012.02602.x

Machlin, L., Miller, A. B., Snyder, J., McLaughlin, K. A.,& Sheridan, M. A. (2019). Differential associations ofdeprivation and threat with cognitive control and fearconditioning in early childhood. Frontiers in BehavioralNeuroscience, 13. https://doi.org/10.3389/fnbeh.2019.00080.

McLaughlin, K. A. (2016). Future directions in childhoodadversity and youth psychopathology. Journal of ClinicalChild and Adolescent Psychology, 45, 361–382. https://doi.org/10.1080/15374416.2015.1110823

McLaughlin, K. A., Greif Green, J., Gruber, M. J., Samp-son, N. A., Zaslavsky, A. M., & Kessler, R. C. (2012).Childhood adversities and first onset of psychiatric dis-orders in a national sample of US adolescents. Archivesof General Psychiatry, 69, 1151–1160. https://doi.org/10.1001/archgenpsychiatry.2011.2277

McLaughlin, K. A., Sheridan, M. A., & Lambert, H. K.(2014). Childhood adversity and neural development:Deprivation and threat as distinct dimensions of earlyexperience. Neuroscience and Biobehavioral Reviews., 47,578–591. https://doi.org/10.1016/j.neubiorev.2014.10.012

McLaughlin, K. A., Sheridan, M. A., & Nelson, C. A.(2017). Neglect as a violation of species-expectant expe-rience: Neurodevelopmental consequences. Biological

16 Rosen et al.

Psychiatry, 82, 462–471. https://doi.org/10.1016/j.biopsych.2017.02.1096

Miyake, A., Friedman, N. P., Rettinger, D. A., Shah, P., &Hegarty, M. (2001). How are visuospatial workingmemory, executive functioning, and spatial abilitiesrelated? Journal of Experimental Psychology: General, 130,621–640. https://doi.org/10.1037/0096-3445.130.4.621

Mott, F. L. (2004). The utility of the HOME-SF scale forchild development research in a large national longitu-dinal survey: The national longitudinal survey of youth1979 cohort. Parenting Science and Practice, 4, 2–3.https://doi.org/10.1080/15295192.2004.9681273

Noble, K. G., Houston, S. M., Brito, N. H., Bartsch, H.,Kan, E., Kuperman, J. M., . . . Sowell, E. R. (2015). Fam-ily income, parental education and brain structure inchildren and adolescents. Nature Neuroscience, 18, 773–778. https://doi.org/10.1038/nn.3983

Noble, K. G., McCandliss, B. D., & Farah, M. J. (2007).Socioeconomic gradients predict individual differencesin neurocognitive abilities. Developmental Science, 10,464–80. https://doi.org/10.1111/j.1467-7687.2007.00600.x

Noble, K. G., Norman, M. F., & Farah, M. J. (2005). Neu-rocognitive correlates of socioeconomic status inkindergarten children. Developmental Science, 8(1), 74–87. https://doi.org/10.1111/j.1467-7687.2005.00394.x

Pearce, D. M. (2017). The self-sufficiency standard for Wash-ington state 2017. Seattle, WA: Workforce DevelopmentCouncil of Seattle-King County.

Peterson, I. T., Bates, J. E., & Staples, A. D. (2015). Therole of language ability and self-regulation in the devel-opment of inattentive-hyperactive behavior problems.Development and Psychopathology, 27, 221–237. https://doi.org/10.1017/S0954579414000698

Ram�ırez-Esparza, N., Garc�ıa-Sierra, A., & Kuhl, P. K.(2014). Look who’s talking: Speech style and social con-text in language input to infants are linked to concur-rent and future speech development. DevelopmentalScience, 17, 880–891. https://doi.org/10.1111/desc.12172

Romeo, R. R., Leonard, J. A., Robinson, S. T., West, M.R., Mackey, A. P., Rowe, M. L., & Gabrieli, J. D. E.(2018). Beyond the 30-million-word gap: Children’sconversational exposure is associated with language-re-lated brain function. Psychological Science, 29, 700–710.https://doi.org/10.1177/0956797617742725

Rosen, M. L., Amso, D., & McLaughlin, K. A. (2019). Therole of the visual association cortex in scaffolding pre-frontal cortex development: A novel mechanism linkingsocioeconomic status and executive function. Develop-mental Cognitive Neuroscience, 39, 100699. https://doi.org/10.1016/j.dcn.2019.100699

Rosen, M. L., Sheridan, M. A., Sambrook, K. A., Meltzoff,A. N., & McLaughlin, K. A. (2018). Socioeconomic dis-parities in academic achievement: A multi-modal inves-tigation of neural mechanisms in children andadolescents. NeuroImage, 173, 298–310. https://doi.org/10.1016/j.neuroimage.2018.02.043

Rowe, M. L. (2012). A longitudinal investigation of therole of quantity and quality of child-directed speechvocabulary development. Child Development, 83,1762–1774. https://doi.org/10.1111/j.1467-8624.2012.01805.x

Samuels, W. E., Tournaki, N., Blackman, S., & Zilinski, C.(2016). Executive functioning predicts academicachievement in middle school: A four-year longitudinalstudy. The Journal of Educational Research, 109, 478–490.https://doi.org/10.1080/00220671.2014.979913

Sarsour, K., Sheridan, M., Jutte, D., Nuru-Jeter, A., Hin-shaw, S., & Boyce, W. T. (2011). Family socioeconomicstatus and child executive functions: The roles of lan-guage, home environment, and single parenthood. Jour-nal of the International Neuropsychological Society, 17,120–132. https://doi.org/10.1017/S1355617710001335

Schrank, F. A., Mather, N., & McGrew, K. S. (2014).Woodcock-Johnson IV tests of achievement. Rolling Mead-ows, IL: Riverside.

Sheridan, M. A., & McLaughlin, K. A. (2014). Dimensionsof early experience and neural development: Depriva-tion and threat. Trends in Cognitive Sciences, 18, 580–585. https://doi.org/10.1016/j.tics.2014.09.001

Sheridan, M. A., & McLaughlin, K. A. (2016). Neurobio-logical models of the impact of adversity on education.Current Opinion in Behavioral Sciences, 10, 108–113.https://doi.org/10.1016/j.cobeha.2016.05.013

Sheridan, M. A., Peverill, M., Finn, A., & McLaughlin, K. A.(2017). Dimensions of childhood adversity have distinctassociations with neural systems underlying executivefunctioning. Development and Psychopathology, 29(5), 1777–1794. https://doi.org/10.1017/S0954579417001390

Slopen, N., McLaughlin, K. A., Fox, N. A., Zeanah, C. H.,& Nelson, C. A. (2012). Alterations in neural processingand psychopathology in children raised in institutions.Archives of General Psychiatry, 69, 1022–1030. https://doi.org/10.1001/archgenpsychiatry.2012.444

Tibu, F., Sheridan, M. A., McLaughlin, K. A., Nelson, C.A., Fox, N. A., & Zeanah, C. H. (2016). Disruptions ofworking memory and inhibition mediate the associa-tion between exposure to institutionalization and symp-toms of attention deficit hyperactivity disorder.Psychological Medicine, 46, 529–541. https://doi.org/10.1017/S0033291715002020

Valloton, C., & Ayoub, C. (2011). Use your words: Therole of language in the development of toddler's self-regulation. Early Childhood Research Quarterly, 26, 169–181. https://doi.org/10.1016/j.ecresq.2010.09.002

Weiss, S. M., Meltzoff, A. N., & Marshall, P. J. (2018).Neural measures of anticipatory bodily attention inchildren: Relations with executive function. Developmen-tal Cognitive Neuroscience, 34, 148–158. https://doi.org/10.1016/j.dcn.2018.08.002

Wiebe, S. A., Espy, K. A., & Charak, D. (2008). Using con-firmatory factor analysis to understand executive con-trol in preschool children: I. Latent structure.Developmental Psychology, 44, 575–587. https://doi.org/10.1037/0012-1649.44.2.575

SES, Cognitive Stimulation, and Executive Function 17

Yousafzai, A. K., Obradovi�c, J., Rasheed, M. A., Rizvi, A.,Portilla, X. A., Tirado-Strayer, N., . . . Memon, U.(2016). Effects of responsive stimulation and nutritioninterventions on children’s development and growth atage 4 years in a disadvantaged population in Pakistan:A longitudinal follow-up of a cluster-randomised facto-rial effectiveness trial. The Lancet Global Health, 4, e548–e558. https://doi.org/10.1016/S2214-109X(16)30100-0

Zelazo, P. D. (2006). The Dimensional Change Card Sort(DCCS): A method of assessing executive function inchildren. Nature Protocols, 1, 297–301. https://doi.org/10.1038/nprot.2006.46

Supporting Information

Additional supporting information may be found inthe online version of this article at the publisher’swebsite:

Table S1. Regression Analyses Including Socioe-conomic Status, Violence Exposure, and CognitiveStimulation in the Same Model

Table S2. Sensitivity Analyses Controlling forVerbal Intelligence

18 Rosen et al.