SES is a multifaceted construct encompassing a family's access to both economic and social resources, and has been examined across various indices, including income/income-to-needs ratio, parental occupational prestige/education, and composites of these and other variables that exhibit differential associations with child cognition (Mousavi et al., 2024; Vrantsidis et al., 2020). SES is associated with an array of critical life experiences that are in turn affiliated with cognitive development (e.g., food insecurity, stressful events; see Jensen et al., 2017, for review), and children of lower SES often score lower on a range of EF assessments than their more affluent peers (Lawson et al., 2018). As will be discussed in the following paragraphs, Domain-General and Context-Specific perspectives provide alternative accounts for how SES contributes to EF—however, given understandable challenges in subjecting SES to experimental manipulation (but see Noble et al., 2021, for an intervention of family income), it should be noted that most studies on SES and related outcomes are inherently correlational.

Traditional Domain-General perspectives often consider associations among SES and EF through deficit lens (see Miller-Cotto et al., 2022, for discussion). These emphasize, for instance, that stressors more frequently experienced within low-SES households are associated with elevated cortisol, which impacts maturation and function of brain regions supporting EF (DePasquale et al., 2021; Vliegenthart et al., 2016; Vogel et al., 2016); notably, low-SES preschoolers show reduced activation in these areas during executive processing than their higher-SES peers (Moriguchi and Shinohara, 2019). By contrast, Context-Specific perspectives suggest that children's experiences and socio-cultural backgrounds are associated with the types of processes that they may engage when exercising control. For instance, variations in children's exposure to conceptual knowledge (e.g., shape, color) may mediate associations among SES and children's performance on laboratory EF assessments that involve these dimensions (e.g., Dimensional Change Card Sort; Perone et al., 2015). Further, the unpredictable circumstances often experienced by children of low SES may also inform a belief that use of control is unlikely to be beneficial, potentially encouraging children to seek immediate rewards over long-term gains (Kidd et al., 2013; Michaelson and Munakata, 2016). These beliefs may be especially influential in novel contexts, such as in a laboratory or with an unfamiliar researcher (Moffett et al., 2020), causing measurements to underestimate children's capacities for control, or undermining the potential benefits of EF interventions (Miller-Cotto et al., 2022). Further, children's perceptions about the predictability of their environments may remain somewhat constant over time, potentially contributing to EF's developmental stability (Michaelson and Munakata, 2016).

Thus, rather than framing SES-based differences in EF as deficits, Context-Specific perspectives more closely align with asset-based approaches, suggesting that these patterns reflect adapted responses to children's socio-cultural contexts. These perspectives are consistent with recent “Hidden Talents” approaches, which posit that adversity strengthens specific cognitive abilities that are especially beneficial in volatile environments (e.g., identifying angry faces; Frankenhuis et al., 2020), and are only revealed in relevant contexts. For example, youth with long-term exposure to adversity (e.g., violence, poverty) tend to perform lower on abstract updating tasks than their non-exposed peers; however, modifying these assessments with ecologically relevant stimuli (e.g., images of emotional faces or money) effectively closes these performance gaps (Young et al., 2022). Such findings offer a valuable reframing of the aforementioned role of SES in cortical development, suggesting that neural connections are developed to suit children's survival needs and other context-specific functions.

As discussed previously, early EF is associated with a variety of cognitive and socio-emotional outcomes (Stucke and Doebel, 2024). EF's associations with numerous facets of academic success (e.g., school readiness, reading/mathematics; see Zelazo et al., 2016, for review), including long-term academic outcomes (Robson et al., 2020), are of particular interest in this review. Many interpret these correlations to imply mechanistic and bidirectional associations among EF and academics, with EF supporting schooling and vice versa (Miller-Cotto and Byrnes, 2020; Peng and Kievit, 2020). Consistent with this perspective, examinations of children of similar ages (5–6 years) but in different grades found that children with one more year of schooling exhibited heightened EF and activation of associated neural substrates relative to their similar-aged peers in lower grades (Brod et al., 2017).

The models highlighted in this review offer different explanations for associations among EF and academic achievement. For example, Domain-General perspectives suggest that schooling provides children with a variety of activities (e.g., circle time; McClelland et al., 2019) that support their broad executive capabilities. These in turn may support a variety of academic tasks, such as paying attention, following instructions, and shifting between activities, perspectives, or strategies (e.g., Chan and Scalise, 2022; Zelazo et al., 2016). Contrasting with generalized approaches, however, various executive processes exhibit differential associations with elements of academic success. For instance, performance on a verbal inhibition task (phonemic Continuous Performance Task) more strongly predicted kindergarteners' early reading than an equivalent non-verbal measure (Foy and Mann, 2013; see also Peng et al., 2024, for similar patterns with elementary school children).

While Domain-General theories may attribute these discrepancies to task impurity, Context-Specific accounts view them as meaningful, suggesting that children engage different processes depending on the specific demands of each academic task. As Context-Specific models characterize EF development as involving the maturation of its constituent processes (e.g., content knowledge; Doebel, 2020), they posit that education directly supports EF by bolstering children's academic skills (Perone et al., 2015). In this view, schooling also provides children with opportunities to practice control pursuant to academic goals (e.g., updating textual information in memory), enhancing their EF in similar educational contexts.

While the theories contrasted in this review offer different explanations for the relationship between EF and academic achievement, both present at least somewhat causal accounts. As with the previously discussed associations among SES and EF, however, most evidence connecting EF and academic achievement is correlational. EF interventions may therefore be especially valuable in elucidating the nature of these processes' associations with academic and other outcomes (Prager et al., 2023).

Numerous studies have linked SES to academic success (Liu et al., 2022), and these associations play out through multiple pathways. For instance, children's socio-economic backgrounds are often directly tied to their school environments; as public schools largely rely on local property taxes for funding, schools within low-SES neighborhoods tend to have limited access to educational resources and offer salaries that are less likely to attract and retain experienced educators (American Psychological Association, 2017; García and Han, 2022). Associations among SES and academic achievement are also mediated by various characteristics of children's home environments, such as familial values regarding education, parental involvement in their children's schooling, and access to educational resources (Bischoff and Owens, 2019; Davis-Kean et al., 2021; Doepke et al., 2019). Of particular interest to this review, associations among SES and academic achievement are partially mediated by EF across the school years (e.g., Lawson and Farah, 2017; Waters et al., 2021).

Given that EF is among many variables linking SES to academic success, it is perhaps unsurprising that the effect sizes of these mediations are often quite modest (e.g., Lawson et al., 2018). However, Context-Specific models are particularly well-suited to account for many of these additional mediating factors. For instance, access to educational resources supports the development of the content knowledge and skills frequently highlighted within these models (e.g., Doebel, 2020; Perone et al., 2021). Similarly, social factors such as familial, peer, and cultural attitudes toward education may also inform students' own values and motivation to succeed academically, which in turn contribute to whether and how students invoke control in academic settings (Miller-Cotto et al., 2022). Rather than viewing these and other mediators as extraneous, Context-Specific models are designed to integrate these variables into a more cohesive theoretical framework. The subsections below discuss the role of several of these factors within Context-Specific models in greater detail.

SES-related gaps in school readiness suggest that differences in key Context-Specific processes emerge well before formal schooling begins. For example, children from families on welfare hear an estimated 30 million fewer words by their 3rd birthdays than their more affluent peers (Hart and Risley, 1995; Ellwood-Lowe et al., 2022; cf. Sperry et al., 2019), with early language gaps predicting later academic success (Lurie et al., 2021). Longitudinal work has documented similar associations among SES and 3-year-olds' mathematical abilities that are primarily mediated by mothers' support of their children's numerical skills (Lombardi and Dearing, 2021). As Context-Specific perspectives frame EF as the assembly of task-specific processes, these views suggest that early gaps in academic skills inherently limit control in contexts where those components are needed (Perone et al., 2021). For instance, children's reading comprehension involves updating textual information in memory, but also foundational skills (e.g., decoding) and content knowledge (e.g., the meaning of words; Artuso and Palladino, 2022) that may vary across socio-economic strata.

These issues may be especially relevant in mathematics, which relies on a wide variety of skills and content knowledge (e.g., number symbols, magnitudes, equations). As the appropriateness of these processes vary considerably across problems, children must assemble the needed processes while inhibiting irrelevant knowledge and strategies (Medrano and Prather, 2023)—thus, basic numerical skills bridge the connection between early EF and mathematical achievement. Associations among Head Start students' EF and performance on a standardized math test were mediated by basic numerical abilities; when all numeracy skills were tested simultaneously, these associations were mediated by a broader range of skills in kindergarteners (set counting, number line estimation, number identification and comparison) than in preschoolers (Chan and Scalise, 2022). Consistent with Context-Specific models' conceptualization of EF development, these findings may suggest that schooling expands and strengthens children's repertoires of numeric and verbal skills, which students may call upon when invoking EF pursuant to academic goals. Whereas these skills are often viewed as sources of statistical noise in Domain-General perspectives, Context-Specific models posit that EF is, in part, the activation of these and other goal-relevant processes (Doebel, 2020).

In addition to goal-specific skills and knowledge, Context-Specific perspectives also emphasize how EF is shaped by socio-contextual factors, such as group norms and expectations. Within this framework, EF is viewed as being informed by children's understanding of what behaviors are expected or valued within their social context, rather than as relatively context-independent traits (Miller-Cotto et al., 2022; Munakata and Michaelson, 2021). In line with this view, preschool-aged children arbitrarily assigned to peer groups (e.g., the “green group”) waited significantly longer on a delay-of-gratification assessment when they believed that other members of their group had been successful at this task (Doebel and Munakata, 2018; Munakata et al., 2020).

These findings have significant implications for children's EF in real-world settings, where their beliefs about their in-groups reflect long-term interactions with their peers and authority figures. Group norms are especially relevant in the classroom given that children of similar SES tend to attend the same schools (Mijs and Roe, 2021). As a result, schools serving low-income districts tend to have higher concentrations of students facing challenges with EF and academics (Bottiani et al., 2019), informing students' collective expectations for their peers' self-regulation and achievement. Teachers also play a significant role in developing classroom norms, and therefore teacher biases (e.g., based on familial income, race) may reinforce or establish maladaptive expectations; research has found that interventions to facilitate positive student-teacher interactions benefit preschoolers' and primary school students' EF (Sankalaite et al., 2021). Context-Specific approaches posit that children's interactions with their peers and teachers inform their internalized beliefs about their own capacities for regulation, and whether attempts at regulation will lead to desirable outcomes (e.g., Doebel, 2020).

Given the effort required for successful control (Diamond, 2020), these beliefs are theorized to factor into a cost-benefit analysis in which children weigh the expenditure of cognitive resources against the likelihood that their efforts will be sufficiently rewarded (e.g., peer approval, good grades). As noted earlier, this decision-making process is particularly relevant among children who often face unpredictable circumstances, and may therefore be especially unsure of whether their control will yield long-term benefits (Kidd et al., 2013; Michaelson and Munakata, 2016; Moffett et al., 2020). This uncertainty or unfamiliarity may contribute to a misalignment among children's context-informed goals (e.g., maximizing immediate rewards) and those of a given situation (e.g., delaying gratification; Frick and Chevalier, 2023). This mismatch may lead to what may be perceived as “poor” EF in both classroom and laboratory settings, potentially causing assessments to underestimate children's capabilities. This is especially relevant for children whose cultural values de-emphasize elements that are often central to standardized EF measures (e.g., obedience to adults, following verbal instructions; Gaskins and Alcalá, 2023). In other words, apparent “failures” of EF may instead reflect adapted responses shaped by children's environments and past experiences (Miller-Cotto et al., 2022).

Context-Specific models therefore frame “successful” EF as the appropriate assembly of cognitive processes given a particular setting or goal, rather than one's generalized use of control. Supporting this view, humans, rodents, and birds reared in unpredictable environments have heightened cognitive skills that are especially beneficial in similarly volatile contexts (e.g., faster attentional shifting to threatening stimuli; see Ellis et al., 2017, for review). Further, for low-SES kindergarteners of color, positive classroom behaviors and performance on a non-delay EF assessment (Head-Toes-Knees-Shoulders) were associated with the tendency to seek immediate gratification on a choice delay task (Duran and Grissmer, 2020). While the authors acknowledged that children's choices may be shaped by other factors that are also linked to SES (e.g., task comprehension, familiarity with “waiting” games; see also Bailey et al., 2021 for further critiques), these findings non-etheless lend support to notion that children's engagement with these tasks are shaped by their socio-cultural contexts.

Though mostly correlational, the findings discussed above point to EF training as a potential means to address SES-based achievement gaps (Lawson and Farah, 2017). EF has vital implications for children's academic skills, as they must, for instance, continually update representations of textual and numerical information in memory. However, although early EF can be trained to an extent, these benefits have unreliably transferred to academic outcomes and other untrained skills (Birtwistle et al., 2025; Melby-Lervåg and Hulme, 2013; Peng and Swanson, 2022; Scionti et al., 2020; Takacs and Kassai, 2019).

Context-Specific models offer insights that may help strengthen the efficacy of EF-based academic interventions. In particular, these perspectives suggest that this lack of transfer is due to a contextual mismatch between the training conditions and their intended outcomes. For instance, training 4-year-old children on traditional EF tasks (e.g., Flanker, Go/No-Go) does not improve their mathematical ability (Blakey et al., 2020), likely because the processes involved in completing these tasks differ significantly from those used in mathematics. These models propose that interventions that train children's EF within specific academic contexts (e.g., training children to update numerical information; Kroesbergen et al., 2012) may be more effective. Consistent with this perspective, a recent meta-analysis on reading interventions in 1st grade and beyond demonstrated that reading-specific EF training programs yielded greater academic improvements than domain-general ones. However, the authors noted that the current pool of studies on context-specific interventions was “underdeveloped,” warranting further research on the effectiveness of these programs (Cartwright and Palian, 2024).

Table 1 overviews studies examining context-specific training programs, focusing on relatively small-scale early childhood interventions targeting academic outcomes—though several large-scale EF interventions (e.g., Tools of the Mind) have yielded impressive academic benefits, they require significant curricular overhaul (and therefore extensive professional development; Diamond et al., 2019) that may be inaccessible to low-SES districts. Consistent with Cartwright and Palian's (2024) assessment, the literature examining context-specific early childhood interventions is also quite limited. The implications of these studies and directions for future research in this area are discussed below.

As seen in the table, a common context-specific intervention strategy is to embed academic content (e.g., numbers, letters) into EF training (or vice versa). This method theoretically benefits academics by having children practice control in similar ways they would use it toward academic pursuits. For example, in a numeracy-modified Dimensional Change Card Sort task, children shifted between counting items based on shape and color (e.g., “how many blue ones are there?”). This training yielded significant improvements in preschoolers' numeracy and general mathematical abilities (Prager et al., 2023). Similarly, training children to flexibly shift between vowel pronunciations bolstered their ability to read and define irregular words (Dyson et al., 2017). While these training interventions generally (but not always; e.g., Kyttälä et al.'s, 2015) improved children's academic outcomes, studies including active control groups demonstrated limited benefits of these programs above and beyond the effects of simply training academic skills in isolation (e.g., Kroesbergen et al., 2012). Further, the few early childhood studies directly comparing context-specific interventions with otherwise equivalent domain-general ones found only small differences in their effects on academic outcomes, though these typically favored context-specific approaches (Kroesbergen et al., 2014; McClelland et al., 2019).

It is entirely possible that EF interventions of any kind will have limited impacts on young children's academic skills. At the same time, these mixed findings suggest that context-specific interventions have potential, and future work can examine modifications of these programs that may boost their effectiveness. For instance, existing interventions can be adjusted to be more explicitly “context-specific”—though often framed as a dichotomy, context-specificity truly exists on a spectrum, with some interventions and measures being more context-specific than others. For example, studies using Kroesbergen et al. (2012) training procedures were included in this review of context-specific interventions because they incorporated numeracy content. However, these may be less context-specific than, for instance, the counting-only training provided to Kyttälä et al. (2015) comparison group. It is likely for this reason that studies using this numbers-based EF training have varied in their classification of this intervention, with one labeling it as “domain-specific” (Kroesbergen et al., 2014), and another as “domain-general” (Kyttälä et al.'s, 2015).

In other words, while these studies included academic content in EF training (and were therefore classified as “context-specific” in this review), there may still be a significant mismatch among training and testing conditions. For instance, these interventions were presented as games, whereas outcomes were often assessed via standardized math and literacy tests (Kroesbergen et al., 2014; McClelland et al., 2019). Though gamification has significant practical benefits for children's engagement and enjoyment of these tasks (Eng et al., 2024), Context-Specific models may caution that this can hinder the contextual alignment between interventions and intended outcomes. Role-playing games in which children apply academic and control skills in semi-realistic scenarios could strike a balance between ecological validity and child engagement. This may help explain the success of programs like Tools of the Mind (e.g., Diamond et al., 2019), though further research is needed to determine whether this approach can be effectively implemented on a smaller, more accessible scale.

In addition to modifying training conditions, this work may also opt for more ecologically valid assessments of academic outcomes in addition to standardized testing. These measures may better capture the effects of context-specific interventions, while also better aligning with these theoretical models. Future studies should also continue to include active control conditions with the goal of developing interventions that substantially benefit academic skills beyond the effects of simply training domain-general EF or academic skills independently.

This line of inquiry could inform interventions to better support EF and academic achievement across a broad range of socio-economic strata. Many of the interventions highlighted in this section may already be relatively accessible to a wide variety of school districts. Most of these studies utilized group training sessions, which may be more easily adapted for a classroom than one-on-one interventions. These training sessions are also relatively short, involve few material resources, and require brief training for faculty/staff planning to implement the intervention (e.g., less than a day; McClelland et al., 2019)—especially compared to more extensive programs, like Tools of the Mind, which requires several days of teacher training (Diamond et al., 2019). At the same time, it is important to note that the faculty/staff, particularly those serving low-SES districts, may already be stretched thin (Farahmandpour and Voelkel, 2025) and any interventions, no matter how “budget-friendly,” may be difficult to implement due to their need for teacher time and cognitive resources—optimizing the academic impact of these interventions will inherently enhance their resource-effectiveness. As discussed above, Context-Specific perspectives may offer a valuable framework for maximizing these programs' benefits.

Context-Specific models broaden our understanding of EF by emphasizing its integration with other cognitive and affective processes, as well as their highly context-dependent nature (Doebel, 2020; Munakata and Michaelson, 2021; Perone et al., 2021). These perspectives are therefore well-suited to provide meaningful insights on how children's environments, including socio-economic factors, shape EF development and related outcomes. As SES is an exceptionally broad construct (Lawson et al., 2018), its linkages with EF and academic outcomes are multi-directional and nearly limitless. Therefore, rather than providing an exhaustive account for these associations, the goal of this review was to provide a conceptual framework that supports future research in exploring SES-related factors not thoroughly addressed here (e.g., food insecurity) and potential interventions to support children's EF and their academic achievement. Further, given that these models underscore the importance of context on EF and related outcomes, additional research may also expand this line of inquiry to the many other interacting systems within children's environments (Bronfenbrenner, 1977). For instance, child cognitive outcomes also reflect dynamics within their households (Davis-Kean et al., 2021), which interact with larger exo- and chrono-system level influences (e.g., systemic racism).

As highlighted previously, a further limitation of this review is that most work on SES and related risk factors/outcomes is inherently correlational, making some of the theorized connections among SES, EF, and academic achievement somewhat speculative. While the academic benefits of the aforementioned EF training paradigms (e.g., Prager et al., 2023) suggest at least somewhat causal associations among these variables, this literature has produced limited findings that warrant further examinations on the effectiveness of these programs. Longitudinal work would also be valuable in examining the long-term effects of these interventions, while also capturing potential developmental shifts in EF's contributions to academic achievement (and vice versa) throughout the school years.

In sum, Context-Specific models provide valuable insights on the interplays among children's environments and cognitive outcomes. This renewed focus is particularly relevant amid widening income disparities (Bischoff and Owens, 2019; Mijs and Roe, 2021)—particularly in the United States, where potential educational reforms may further disadvantage children and teachers within low-income districts. Traditional and emerging models each contribute to our understanding of how EF and its constituent processes can be supported in low-SES and other at-risk children (Doebel, 2020; Ibbotson, 2023; Zelazo and Carlson, 2023). This work is especially important given that a recent meta-analysis noted a severe lack of attention to SES in the developmental sciences (Singh and Rajendra, 2024). Continued efforts in this area will not only advance our scientific understanding of EF and its development, but also support vital cognitive and academic outcomes.