In the developmental trajectory of human cognition, few phenomena illustrate the complex orchestration of perception, action, and executive control quite as vividly as the scale error. A twenty-month-old toddler, having spent several minutes joyfully sliding down an indoor playground slide, casually turns toward a dollhouse replica of the exact same slide—measuring no more than a few inches in length—and with absolute seriousness attempts to place their foot onto the diminutive plastic ladder. In another corner of the laboratory, an otherwise typically developing child attempts to wedge their entire body into a miniature toy car the size of a shoebox, pressing their shoe against the tiny door with evident force and perplexity when the vehicle fails to accommodate them. These bizarre, striking, and momentarily baffling behaviors are not acts of whimsical make-believe, nor are they products of clinical visual impairment. Instead, they represent a profound, transient window into the modular architecture of the developing human brain, known universally in contemporary developmental psychology as the scale error.
First formally isolated and systematically investigated by developmental psychologist Judy DeLoache and her colleagues in a landmark 2004 study published in the journal Science, the scale error paradigm reshaped foundational assumptions regarding how young children translate visual perception into motor planning. For decades, traditional Piagetian frameworks had posited that perception and action in early childhood operate in relative sensorimotor synchrony, gradually refined through experiential assimilation and accommodation. DeLoache’s empirical discoveries shattered this monolithic view, providing compelling behavioral evidence that visual perception in toddlers is fundamentally segregated into distinct, asynchronously maturing neurocognitive processing streams. A toddler can simultaneously identify what an object is while completely miscalculating how their physical body can interface with it in three-dimensional space.
To fully appreciate the significance of DeLoache’s scale error experiment, one must examine the intricate intersections of evolutionary biology, developmental cognitive neuroscience, ecological psychology, and executive function. The scale error is not merely an amusing quirk of toddlerhood; it is an evolutionary artifact of a cognitive system under rapid structural remodeling. By investigating why children between the ages of 18 and 30 months attempt the physically impossible, cognitive scientists have unraveled vital clues concerning the bifurcated visual pathways of the brain, the protracted maturation of the prefrontal cortex, the mechanics of affordance perception, and the elusive nature of dual representation. The following comprehensive exploration deconstructs the scale error phenomenon from its historical and theoretical roots to its modern legacy in developmental neuroscience and artificial intelligence.
1. Historical and Theoretical Antecedents of DeLoache’s Research
1.1 The Evolution of Toddler Cognitive Architecture Models
To conceptualize the emergence of the scale error paradigm, one must trace the theoretical evolution of toddler cognitive architecture across the twentieth century. For decades, the dominant paradigm governing early cognitive development was the sensorimotor framework advanced by Jean Piaget. In Piaget’s constructivist epistemology, the infant and young toddler construct mental representations through active, physical manipulation of the environment. Piaget posited that between 18 and 24 months of age, children transition from the sixth stage of sensorimotor intelligence—characterized by the invention of new means through mental combinations—into the preoperational stage, marked by the dawn of symbolic thought and internal representation. Crucially, classical Piagetian theory assumed that perceptual recognition and motor interaction were inextricably bound; an action schema was deemed to be directly calibrated by the child’s sensorimotor experience with physical objects. If a toddler understood the functional category of a chair, that knowledge was presumed to be co-extensive with their physical interaction with chairs.
In stark contrast to Piaget’s constructivism stood the ecological psychology of James J. Gibson and Eleanor J. Gibson. The Gibsonian framework dispensed with the necessity of complex mental representations, arguing instead for direct perception. Central to Gibson’s theory was the concept of an affordance: the actionable properties between an organism and its physical environment. According to Gibson, an affordance is not an abstract cognitive category but an objectively real, directly perceived relational property. A flat, rigid, horizontal surface of an appropriate height relative to an individual’s knee joint directly affords “sitting-on.” For Eleanor Gibson, developmental maturation involved the progressive education of attention—infants and toddlers learned to detect increasingly subtle visual information specifying whether an environmental feature afforded a given motor act, such as walking down a steep slope or crawling across an apparent visual cliff. Gibsonian theory inherently presumed that healthy visual perception automatically coupled environmental scale to the actor’s bodily dimensions.
Concurrently, cognitive developmentalists during the late twentieth century began dismantling the idea that infants were entirely devoid of representational competence prior to the preoperational stage. Pioneering studies in infant cognition revealed that even young babies possessed rich, innate or early-emerging core knowledge systems regarding physical causality, object permanence, and solidity. However, a persistent theoretical rift remained regarding how symbolic comprehension emerged from these basic cognitive competencies. How did a child transition from treating an object purely as a physical entity to treating it as a symbol that stands for something else? While infants could perceive objects, and older children could navigate symbolic systems, the intervening period—the toddler years—remained a theoretical battleground characterized by fragile, uneven representational competence where bodily action and mental representation frequently collided.
1.2 Judy DeLoache’s Prior Work on Dual Representation
Before unveiling the scale error experiment, Judy DeLoache had already fundamentally transformed developmental psychology through her pioneering research on symbolic understanding and the formulation of the Dual Representation Hypothesis. Throughout the late 1980s and 1990s, DeLoache investigated how young children comprehend physical scale models. In her iconic “model room” experiments, toddlers were introduced to a full-sized room containing standard furniture (a couch, a chair, a floor lamp, a rug) and a precisely scaled-down, miniature replica of that exact same room. An experimenter would hide a miniature toy (such as a tiny plastic dog) behind a piece of miniature furniture in the scale model, while the child observed. The child was then escorted into the real, full-sized room and asked to find a corresponding larger toy (a large stuffed dog) hidden in the exact analogous location.
DeLoache’s findings were striking and robust: children at 36 months of age systematically succeeded at this retrieval task, immediately using the spatial information observed in the miniature model to locate the object in the full-sized room. Conversely, children at 30 months of age failed completely. They remembered where the miniature toy was hidden in the model room, but they could not translate that spatial knowledge to the real room. To explain this stark developmental divide, DeLoache articulated the dual representation hypothesis. For a child to use a physical scale model as a symbolic tool, they must achieve a dual representation: they must mentally represent the model simultaneously as a physical, tangible object in its own right (with its own concrete properties, colors, and textures) and as a symbol representing a different, distal physical reality.
DeLoache demonstrated that this dual cognitive demand placed an immense burden on the toddler’s immature cognitive architecture. Because the scale model was an intrinsically fascinating, highly salient physical object filled with miniature chairs and tiny tables, the 2.5-year-old child became mentally fixated on its concrete reality. The physical salience of the object actively obscured its symbolic function. To validate this hypothesis, DeLoache famously executed the “shrinking room” experiment, in which toddlers were led to believe that an elaborate “shrinking machine” had literally transformed the full-sized room into the miniature model. In this condition, where the child believed the model was the room rather than a symbol for the room, dual representation was no longer required; consequently, even 30-month-old toddlers succeeded effortlessly in locating the toy. This profound line of inquiry revealed that toddlers experience severe cognitive vulnerability whenever they are forced to negotiate the physical reality of an artifact alongside its conceptual meaning. It laid the direct conceptual foundation for shifting focus from symbolic retrieval tasks to direct, unmediated motoric interactions with miniaturized objects.
1.3 Emergence of the Scale Error Paradigm
The genesis of the scale error paradigm was not the product of a purely deductive theoretical exercise, but rather a classic instance of scientific serendipity intersecting with prepared, rigorous observational acumen. While conducting research on symbolic comprehension and spatial cognition in her developmental laboratory, DeLoache and her research team repeatedly observed anomalous, unprompted behaviors exhibited by toddler participants during unstructured transition periods. Children who had finished an experimental protocol would occasionally walk over to miniature play artifacts scattered in the laboratory—artifacts intended purely as props or symbolic models—and attempt to interact with them as if they were full-scale functional entities.
In one memorable laboratory anecdote that crystallized the inquiry, a young toddler finished playing in a play area with a child-sized plastic automobile. After being briefly distracted, the child walked past a miniature die-cast toy car resting on the floor, crouched down, and earnestly attempted to insert his foot through the quarter-inch driver-side opening, pushing down with an earnest, focused demeanor until he lost his balance. In another instance, a toddler girl meticulously tried to sit her full body down onto an ornate dollhouse chair, despite the chair being shorter than her ankle. While casual observers might have dismissed these bizarre episodes as whimsical pretend play or momentary toddler silliness, DeLoache recognized that the motor kinematics, facial affect, and physical persistence of the children pointed toward something far more fundamental: a catastrophic breakdown in the visual-motor control loop.
DeLoache recognized the critical imperative to distinguish these bizarre incidents from genuine pretend play. In symbolic play, toddlers typically exhibit exaggerated vocalizations, smiling, shared gaze with caregivers, and delicate, pantomimed physical contact; they do not apply full body weight to fragile miniatures. The children in DeLoache’s laboratory, by contrast, were applying genuine mechanical force, exhibiting serious, neutral, or perplexed facial expressions, and appearing genuinely startled when their bodies failed to interface with the miniature objects. Recognizing that this phenomenon represented an unprecedented window into the segregation of action planning and size perception, DeLoache, alongside colleagues David H. Uttal and Karl S. Rosengren, formulated the standardized empirical paradigm that culminated in their historic 2004 paper published in Science, entitled “Scale Errors Offer Evidence for a Perception-Action Dissociation Early in Life.”
2. Defining the Scale Error Phenomenon in Early Childhood
2.1 Operational Definition and Core Characteristics
In developmental cognitive science, a scale error is operationally defined as a serious, unprompted attempt by a young child to perform an action on a miniature replica of an object that is physically impossible to execute due to the extreme discrepancy between the child’s physical body dimensions and the dimensions of the target object. This formal definition hinges critically on the operationalization of “serious attempt.” A behavior is coded as a genuine scale error only when the child demonstrates unequivocal motoric commitment to the functional action schema traditionally associated with the full-sized referent object, rather than a symbolic, metaphorical, or playful pantomime of that action.
The behavioral coding criteria established by DeLoache and colleagues require that the action plan must be directed toward the miniature object with the standard biomechanical trajectory and force parameters typical of the actual behavior. For example, if a child encounters a miniature replica of an armchair that measures roughly 10 centimeters in height, a coded scale error requires that the child turns their back to the object, bends their knees, and actively lowers their full body weight down with the clear motor objective of resting their buttocks on the diminutive surface. The interaction cannot simply be a casual, exploratory tap with the hand or an affectionate patting of the toy. The child must genuinely treat the miniature object as an actionable entity for their own body.
Another definitive hallmark of the scale error is its transient, age-delimited nature. The phenomenon is uniquely concentrated within a specific developmental window, appearing almost exclusively between the ages of 18 and 30 months, with a peak incidence occurring between 20 and 24 months. Prior to 18 months, while infants certainly display a range of motor exploratory behaviors, they rarely possess the stable, automated motor action schemas necessary to mount an organized, complex behavioral sequence (such as systematically climbing into a vehicle) toward an incongruent object. After 30 months of age, the phenomenon dissipates rapidly; typically developing children over the age of two and a half almost never commit scale errors under standard environmental conditions, instead interacting with miniature objects through explicitly symbolic, verbal, or fine-motor play schemas.
2.2 Taxonomy of Common Scale Errors
Empirical investigations across numerous laboratories have established a rigorous taxonomy of common scale errors observed in early childhood, categorized primarily by the target functional artifact and the corresponding bodily action schema invoked. The most iconic and widely documented category involves postural seating errors directed toward miniature furniture. In these instances, toddlers encounter miniature chairs, stools, or benches—ranging from sturdy dollhouse furnishings to foot-high decorative models—and attempt to execute complete sitting maneuvers. The child systematically positions their pelvic region over the miniature target, frequently using their hands to grasp the miniature armrests or edges, and lowers themselves completely, frequently resulting in a loss of balance, falling backward onto the floor, or awkwardly straddling the microscopic seat while remaining supported solely by their own flexed leg musculature.
A second prominent category encompasses locomotor transit errors involving miniature sliding apparatuses. Toddlers exposed to diminutive toy slides (often manufactured for miniature dolls or action figures, spanning less than twelve inches in total length) attempt to climb the tiny structural ladder or clamber directly onto the microscopic slide platform. The child will earnestly attempt to place their foot onto rungs that are narrower than their own toes, grasp the tiny handrails with a pincer-like grip, and attempt to slide their full torso down the precipitous plastic incline. Such attempts are characterized by rigorous, forward-directed biomechanical momentum, often halted only when the child’s knee or pelvis physically collides with the floor surrounding the miniature apparatus.
A third major taxonomic grouping consists of vehicular insertion errors. When confronted with miniature toy cars, wagons, trucks, or ride-on toys that are dramatically smaller than their body—frequently measuring only 15 to 25 centimeters in length—toddlers will systematically attempt to climb inside. This behavior typically begins with the child opening the tiny vehicle door (if functional) or orienting their body directly over the miniature cabin, followed by an earnest, forceful effort to plant one foot directly into the tiny interior cabin. The child will repeatedly press downward, attempting to force their foot into a compartment that could barely accommodate a walnut, often leaning their entire body mass over the toy in a persistent effort to achieve entry.
Finally, researchers have documented a broad array of fine-motor and vehicular accessory errors. These include toddlers attempting to push their feet into miniature plastic doll shoes, trying to wrap miniature bracelets designed for dolls around their own wrists, or applying their full hand to microscopic door handles, keys, and cabinet latches that are sub-millimeter in scale. In each of these diverse taxonomic expressions, the structural constant remains identical: an overarching motor program historically perfected on a full-scale environmental object is inappropriately mobilized and directed toward a geometrically identical object of radically inadequate physical scale.
2.3 Distinction from Other Early Childhood Motor Errors
To establish the scientific validity of the scale error as an independent neurocognitive phenomenon, researchers have taken great care to differentiate it from other common motoric, perceptual, and executive errors that characterize early childhood. Most fundamentally, a scale error must be rigorously distinguished from general motor clumsiness or coordination deficits. Toddlers are notoriously unstable bipeds; their gait is wide-based, their center of mass is high, and their balance control mechanisms are undergoing continuous calibration. However, general motor clumsiness involves the faulty execution of an otherwise appropriate motor plan—such as tripping over an obstacle, misjudging a step height by a few centimeters, or failing to grasp a cup securely due to inadequate grip force calibration. In a scale error, the motor execution itself is often mechanically flawless in its kinematic form; the pathology lies entirely in the catastrophic contextual inappropriateness of executing that specific motor program upon an object of that specific physical metric.
Furthermore, scale errors are conceptually distinct from perceptual classification errors or conceptual category confusion. If a young child points to a miniature plastic cow and calls it a “horsey,” or mistakes a small spherical decorative marble for an edible piece of candy, the child is committing a semantic or perceptual categorization error based on shared physical attributes. In a scale error, the child does not misidentify the category of the object; on the contrary, the child identifies the category with hyper-precision! The toddler recognizes that the miniature car belongs to the category “car,” and that cars are entities meant to be sat in and driven. The error does not stem from a failure to identify what the object represents, but rather from a failure to integrate the object’s real-time, physical dimensions into the motor planning apparatus.
Finally, scale errors must be differentiated from classic perseverative motor errors, such as the renowned Piagetian A-not-B error observed in younger infants. In the classic A-not-B task, an infant continues to reach for an object at location A where it was previously found, even after watching it be hidden at location B; this represents an executive failure to inhibit a previously rewarded motor habit in the presence of conflicting spatial evidence. While scale errors share an element of inhibitory failure, they do not require a historical conditioning loop of repetitive physical reinforcement at a specific spatial coordinate immediately prior to the error. Scale errors can occur instantaneously upon the first visual presentation of a miniature stimulus, driven purely by the visual recognition of the object’s form, entirely independent of repetitive spatial search conditioning.
3. The Classic 2004 DeLoache Paradigm: Experimental Methodology
3.1 Participant Cohort and Developmental Staging
The foundational experimental methodology designed by Judy DeLoache, David Uttal, and Karl Rosengren, published in their 2004 Science paper, was engineered to capture spontaneous, naturalistic motor behaviors under rigorously controlled laboratory conditions. The participant cohort comprised 54 typically developing toddlers, meticulously stratified across the critical developmental window spanning 18 to 30 months of age. To assess potential developmental trajectories and age-dependent attenuation, the cohort was divided into distinct, cross-sectional age bands: younger toddlers (averaging approximately 20 to 22 months of age) and older toddlers (averaging 26 to 28 months of age). Both male and female toddlers were represented across the cohorts to ensure that gender-specific play socialization did not confound the experimental findings.
Inclusion criteria required that all participating children were born full-term, possessed normal or corrected-to-normal visual acuity, had reached standard motor milestones (specifically, independent bipedal locomotion for at least four months), and had no documented developmental delays or neurodevelopmental disorders. Prior to testing, parents completed standardized baseline developmental screening questionnaires, including the MacArthur-Bates Communicative Development Inventories (CDI) to gauge receptive and expressive vocabulary, as well as behavioral questionnaires assessing the child’s daily familiarity with standard domestic furnishings and toy vehicles. The experimental sessions were conducted in a child-friendly, naturalistic laboratory play space designed to minimize separation anxiety while maximizing observational fidelity.
3.2 Apparatus and Physical Stimuli Design
The core physical apparatus developed for the 2004 paradigm consisted of pairs of precisely calibrated physical stimuli: three full-sized, fully functional children’s play objects, and their three exact miniature replica counterparts. The three functional objects were selected because they represented familiar, highly salient artifacts that elicit distinct, stereotyped whole-body motor schemas in young toddlers: a child-sized upholstered armchair, an indoor plastic playground slide with a ladder, and a large, durable plastic ride-on toy car with an openable driver’s door and a functional interior seat.
The engineering of the miniature counterparts was central to the paradigm’s internal validity. Rather than utilizing generic, disparate miniature toys, DeLoache and her colleagues manufactured or procured miniature replicas that matched their full-scale counterparts with absolute geometric fidelity. The miniature armchair, the miniature slide, and the miniature car were identical to the large objects in color, textural material, surface finish, structural proportions, and distinct decorative markings; they were, in every sensory dimension except absolute volume, perfect scaled-down models. The miniature armchair measured approximately 10 to 12 centimeters in height; the slide measured roughly 15 centimeters in total height; and the miniature car measured approximately 18 centimeters in length. The spatial scaling factor was approximately 1:6 to 1:8 relative to the large objects. This rigorous geometric correspondence ensured that any behavioral activation observed could be attributed directly to the shared perceptual features of the objects rather than confounding visual discrepancies.
3.3 Standardized Experimental Protocol and Procedure
The experimental procedure followed a standardized, multi-phase protocol structured into four continuous stages: Familiarization, Distraction, Substitution, and Observation. In the Familiarization Phase, the toddler entered the laboratory playroom containing the three full-sized, functional objects. The experimenter and the child’s parent actively encouraged the child to interact freely with each object. The child sat in the large armchair, climbed up the steps and slid down the full-sized slide, and climbed inside the large toy car to push themselves along the floor. This familiarization phase lasted roughly ten minutes, serving a dual scientific purpose: it confirmed that the child possessed the physical competence to execute each motor schema, and it primed the appropriate behavioral affordances associated with each functional object.
Following this phase, the Distraction Interval was introduced. The experimenter gently guided the toddler out of the primary testing room, or directed their attention toward an adjacent observation alcove under the pretense of looking at a picture book, engaging with a wall-mounted activity board, or receiving a small sticker. This distraction lasted approximately two to three minutes, ensuring that the child’s visual attention was completely disengaged from the primary play space while keeping their cognitive arousal and comfort levels stable.
During the child’s absence from the primary play area, the surreptitious Stimulus Substitution took place. The research assistants swiftly removed the three full-sized objects from the room and placed the three miniature replica counterparts in the exact spatial locations and physical orientations previously occupied by the large objects. The miniature armchair sat on the precise patch of carpet where the large armchair had rested; the miniature slide mirrored the alignment of the full-scale slide; and the miniature car was oriented with its tiny driver’s door facing the identical direction. This spatial invariance was designed to harness the child’s spatial expectations and environmental memory.
Finally, the Observation Phase commenced. The child was escorted back into the primary testing room. The experimenter adopted a non-directive, neutral stance, neither pointing toward the miniature objects nor issuing verbal commands regarding what to do with them. The experimenter simply remarked neutrally, “You can play with whatever you want.” Continuous audiovisual recording captured every movement of the child from multiple angles. Research personnel tracked the child’s latency to approach each stimulus, the trajectory of their approach, the nature of their initial physical contact, the mechanical force applied, and their expressive behaviors upon encountering the miniature items.
3.4 Video Coding Schema and Reliability Measures
To eliminate observational subjectivity, DeLoache’s laboratory constructed a rigorous, micro-analytic video coding schema. Two independent, highly trained developmental coders, blinded to the specific developmental hypotheses and baseline cognitive scores of the children, conducted frame-by-frame analysis of all recorded sessions. The schema established explicit operational definitions to discriminate genuine scale errors from exploratory manipulation, accidental physical collisions, and symbolic pretend play.
A behavior was coded as a genuine scale error only if the child exhibited full biomechanical commitment to the functional action schema. For the miniature chair, the child had to turn, squat, and attempt to deposit their pelvic weight onto the miniature seat. For the miniature slide, the child had to attempt to position their foot on the ladder rungs or place their body in a sliding posture on the miniature ramp. For the miniature car, the child had to attempt to place their foot inside the tiny cabin, straddle the vehicle, or sit upon the diminutive roof. Coders simultaneously tracked micro-indices of facial affect and gaze: the presence or absence of smiling, laughter, vocalizations of make-believe (e.g., “vroom vroom” or doll-directed speech), and whether the child exhibited signs of perplexity, neutral determination, or gaze-checking toward their parent. If a child picked up the miniature car and rolled it along the carpet, or sat next to the miniature chair and placed a stuffed animal onto it, the behavior was rigorously coded as appropriate miniature toy play or pretense, explicitly excluding it from the scale error count.
Inter-rater reliability was assessed using Cohen’s kappa ($kappa$), yielding high concordance rates routinely exceeding 0.90 across coded categories. Any ambiguous behavioral sequence—such as a toddler momentarily placing a foot near the car without downward force, or brushing against the slide while reaching for another toy—was submitted to a panel review of senior developmental researchers. If a consensus confirming earnest motor intent could not be unanimously reached, the behavior was conservatively discarded from the scale error tally. This stringent methodological conservativism guaranteed that the resulting empirical data represented an incontrovertible baseline of genuine motoric failures.
4. Neurological Mechanisms: The Two Visual Systems Hypothesis
4.1 The Milner and Goodale Dual-Stream Model
To explain the neurobiological etiology of the scale error, Judy DeLoache and her colleagues turned to one of the most influential frameworks in contemporary cognitive neuroscience: the dual-stream model of visual processing formulated by neuroscientists A. David Milner and Melvyn A. Goodale (1995). Departing from the classical Ungerleider and Mishkin “what versus where” dichotomy, Milner and Goodale proposed that cortical visual processing in primates is bifurcated into two anatomically and functionally distinct streams: the ventral stream (the “vision-for-perception” or “what” pathway) and the dorsal stream (the “vision-for-action” or “how” pathway).
The ventral stream originates in the primary visual cortex (V1) and projects forward through visual area V4 into the rich, associative networks of the inferotemporal (IT) cortex. The primary computational objective of the ventral stream is the conscious identification, categorization, and semantic evaluation of visual stimuli. Crucially, the ventral stream operates in an allocentric (object-centered) frame of reference and is characterized by long-term representational memory and perceptual constancy. It recognizes an armchair as an “armchair” regardless of whether that armchair is viewed from the front, from the side, under varying illumination conditions, or—critically—at different visual angles and metric sizes. It identifies the categorical essence and functional identity of an object.
Conversely, the dorsal stream projects from V1 through intermediate visual areas (such as V6 and MT/V5) directly into the posterior parietal cortex (PPC). The computational mandate of the dorsal stream is the real-time, millisecond-by-millisecond control of unreflective, goal-directed physical action. Unlike the ventral stream, the dorsal stream operates strictly in an egocentric (body-centered) coordinate system. It does not concern itself with what an object is called or what abstract semantic category it belongs to; rather, it rapidly computes the exact metric physical properties of the object relative to the actor’s physical effectors—its precise distance, its absolute three-dimensional volume, its surface orientation, and its immediate affordances for grasping, stepping, or sitting. Dorsal computations are ephemeral and non-conscious, decaying almost instantaneously once the visual stimulus is removed, ensuring that motor plans are continuously updated based on the immediate physical reality of the physical environment.
4.2 Failure of Stream Integration as the Primary Etiology
The core theoretical thesis advanced by DeLoache, Uttal, and Rosengren is that a scale error represents a dramatic, momentary failure of functional integration between these two asynchronously developing visual processing streams. When a toddler enters the playroom and sees the miniature replica armchair, the visual sensory input triggers an immediate, automatic cascade of perceptual recognition within the ventral stream. The ventral stream recognizes the distinct shape, color, armrests, and backrest of the object, successfully categorizing it as an instance of the familiar category “armchair.” Concurrently, this ventral identification activates an associated, deeply ingrained behavioral schema stored in the motor cortex and basal ganglia: “armchair = an object to sit upon.”
Under normal, mature neurological functioning in older children and adults, this initial ventral activation is instantaneously cross-referenced, calibrated, and—if necessary—overridden by the dorsal stream. As the individual begins to approach the object and formulate an actual motor command, the posterior parietal cortex (operating via the dorsal pathway) computes the immediate, real-time spatial metrics of the physical target: it registers that the surface is only 10 centimeters off the floor, that its width is entirely insufficient to support the adult pelvis, and that its load-bearing capacity is inadequate. In the mature brain, this dorsal metric assessment immediately suppresses the inappropriate sitting motor plan, substituting an alternative motor program suited to the miniature dimensions (such as bending down to pick it up with the fingers).
In the 20-month-old toddler, however, this cross-stream inhibitory loop fails catastrophically. The ventral stream’s categorical recognition of the object successfully mobilizes the general action plan to sit down. But as the child executes the motor approach, the dorsal stream—despite accurately processing the small visual angle of the object—fails to exert timely, top-down veto power over the already activated ventral action template. There is a temporary decoupling between visual size perception and real-time motor programming. The toddler’s motor system proceeds to execute the ventral action plan using feedforward commands calibrated for a full-sized armchair, completely failing to update the motor effector parameters to reflect the miniature metric reality detected by their visual system.
4.3 Cortical Underpinnings of Size-Affordance Processing
This failure of stream integration is not a structural defect, but a predictable consequence of the protracted, asynchronous structural maturation of the human cerebral cortex. Neuroimaging and post-mortem histological studies reveal that the cortical regions underpinning the ventral stream, particularly the primary visual areas and the ventral occipitotemporal cortex, reach functional and synaptogenic maturity significantly earlier in ontogeny than the complex associative networks of the posterior parietal cortex and its reciprocal projections to the prefrontal cortex.
The posterior parietal cortex (PPC) is the vital neuroanatomical hub responsible for calculating size-affordance interactions. Sub-regions within the intraparietal sulcus (IPS), including the anterior intraparietal area (AIP), contain specialized visuomotor neurons that code for hand shaping, grip aperture, and postural orientation based on the immediate 3D physical geometry of an object. In toddlers aged 18 to 24 months, the functional connectivity between the inferotemporal cortex (the seat of ventral semantic identity) and the posterior parietal cortex is still undergoing extensive white matter myelination. The long-range association fiber tracts—such as the vertical occipital fasciculus and the inferior longitudinal fasciculus—are structurally immature.
Consequently, while the inferotemporal cortex can rapidly transmit categorical identification signals forward to premotor and motor areas, the corresponding parieto-frontal corrective circuits lack the processing speed and synaptic robustness required to override a prepotent motor response in real time. The toddler’s brain experiences a temporal lag: the motor command is already ballistic and descending the corticospinal tract before the posterior parietal cortex can register that the metric affordance is mathematically invalid. This neurological disconnect precisely accounts for why toddlers frequently look directly at the miniature object, appear visually focused, and yet still attempt an action that their own visual system should theoretically register as an anatomical impossibility.
5. The Role of Inhibitory Control and Executive Function
5.1 Immature Prefrontal Inhibition Networks
While the dual-stream dissociation provides the sensory-motor explanation for scale errors, it must be complemented by an analysis of the toddler’s executive function architecture. A scale error is fundamentally an inhibitory control failure. Even if the ventral visual pathway generates an inappropriate action plan based on categorical recognition, a cognitively mature system would rely on executive control mechanisms to halt the behavioral cascade before it achieves motoric execution. This inhibitory veto is the evolutionary responsibility of the prefrontal cortex (PFC).
The dorsolateral prefrontal cortex (DLPFC) and the orbitofrontal cortex are widely recognized as having the most protracted developmental timeline of any human brain regions, with synaptic pruning and axonal myelination continuing well into late adolescence and early adulthood. At 18 to 24 months of age, the DLPFC is profoundly immature. As pioneering developmental neuroscientist Adele Diamond has exhaustively demonstrated in her work on infant executive function, young children exhibit severe limitations in response inhibition whenever a prepotent, highly familiar motor habit is triggered by an emotionally or perceptually salient stimulus.
In the context of the scale error paradigm, the functional action schema associated with an armchair, a slide, or a car represents a massive, over-learned, prepotent motor response. For a toddler, the visual presentation of a chair has been paired hundreds of times with the rewarding motor act of sitting. Once the child’s ventral stream identifies the miniature replica as an armchair, that deeply entrenched sitting schema is instantly primed at an electrophysiological level. Because the child’s prefrontal inhibitory networks are insufficient to exert top-down suppression, the threshold for motor discharge is rapidly breached. The child cannot “hit the brakes” on the motor plan. This executive failure mirrors the cognitive architecture of the A-not-B error: in both cases, an immature frontal-striatal circuit fails to inhibit an automated, prepotent motor output, even though perceptual information signaling the error is readily available to the sensory cortex.
5.2 Action Planning Dynamics and Cascading Motor Output
To fully grasp the biomechanics of the scale error, one must conceptualize toddler action planning through the lens of computational motor control models, specifically feedforward versus feedback control mechanisms. In adult motor execution, action is governed by a tightly integrated combination of predictive feedforward commands and real-time sensory feedback loops. When an adult reaches to grasp an object, the brain initiates a feedforward motor program based on an initial visual estimate of the object’s location and size, but as the hand approaches the target, visual and proprioceptive feedback mechanisms continuously fine-tune the grip aperture and muscular force, terminating the movement gracefully if the object suddenly shifts or proves physically incompatible.
In toddlers between 18 and 30 months, motor planning is notoriously ballistic. Because neural transmission speeds in unmyelinated peripheral and central pathways are substantially slower than in adults, toddlers rely overwhelmingly on open-loop, feedforward motor cascades. Once an intentional motor schema is selected by the premotor cortex, it tends to be deployed as a single, indivisible, non-correctable burst of motor output. When a 20-month-old toddler resolves to sit in the miniature car, the entire neuromuscular sequence—orienting the pelvis, lifting the leg, shifting the center of gravity, and driving the foot downward—is fired off in a ballistic cascade.
Because the toddler’s real-time proprioceptive and visual feedback correction loops are slow and poorly integrated with ongoing motor execution, the child does not process the physical absurdity of the interaction until mechanical resistance is physically encountered. The toddler does not pause mid-reach to re-evaluate; rather, the action plans cascade uncontrollably into physical contact. The child’s downward momentum continues until external physical constraints—such as their shoe colliding solidly with the plastic floorboard of the miniature car, or their buttocks striking the carpet next to the dollhouse chair—force a mechanical cessation of the action.
5.3 Attentional Capture by Visual Similarity
A third critical cognitive mechanism contributing to scale errors is the phenomenon of perceptual attentional capture driven by high visual similarity. The miniature replicas utilized in DeLoache’s standardized paradigms were intentionally crafted to be identical to their full-scale counterparts in every perceptual dimension except absolute spatial metric. They shared identical color hues, reflective gloss, surface textures, geometric shapes, and micro-details such as miniature handles, cushions, and steering wheels.
For the toddler’s visual attention system, which is largely driven by bottom-up, stimulus-driven salience rather than top-down, goal-directed filtering, these hyper-salient visual features exert an irresistible pull. Developmental research indicates that young children exhibit a cognitive bias known as feature-based attentional capture: their attention is disproportionately captured by recognizable sub-features (e.g., the bright red steering wheel or the plush fabric of the seat) rather than by global spatial dimensions (such as the total cubic volume of the artifact). The high visual fidelity of the miniature replica effectively “blinds” the toddler’s attentional spotlight to the global dimensional scale of the object.
This attentional capture creates a severe cognitive load. As the toddler visually fixates on the familiar, highly rewarding visual sub-features of the miniature object, their working memory capacity becomes fully saturated. In an immature cognitive system with severely limited processing bandwidth, the cognitive resources required to simultaneously compute metric size, compare that size to their own body schema, and coordinate an inhibitory veto are simply unavailable. Perceptual salience triumphs over geometric metric assessment, driving the child forward into an impossible physical act.
6. Behavioral Patterns and Observational Typologies of Scale Errors
6.1 Biomechanical Features of Scale Error Manifestations
The behavioral manifestations of scale errors are characterized by specific, highly stereotyped biomechanical signatures that completely differentiate them from ordinary childhood clumsiness or delicate play. When a child commits a scale error, the physical force applied to the miniature object is remarkably substantial. Biomechanical video analysis reveals that toddlers do not gingerly test the structural integrity of the miniature object; instead, they commit a substantial percentage of their total body mass to the functional action schema.
In sitting errors directed toward miniature chairs, children routinely exhibit a full, ballistic backward descent. They bend their knees, extend their arms forward to counterbalance their torso—a standard human postural adjustment for sitting—and lower their center of mass directly over the miniature object. In many observed cases, the child’s downward kinetic energy completely crushes a fragile miniature or causes the sturdy replica to violently shoot out from beneath them, resulting in a sudden loss of balance and the child falling flat onto the floor. Despite this physical collision, the child often does not exhibit immediate pain or fear; instead, they appear physically surprised by the abrupt ground contact.
Similarly, vehicular insertion errors exhibit extraordinary physical force. Toddlers attempting to climb into a 15-centimeter toy car will place their entire foot onto the vehicle, pressing down with sufficient downward force that their quadriceps tense, their calf muscles contract, and their opposite foot leaves the floor, momentarily balancing their entire body weight on top of the miniature toy. When the foot inevitably slips off or fails to penetrate the cabin, the child frequently stumbles, re-stabilizes their posture, and immediately re-attempts the exact same movement. This mechanical persistence—the tendency to repeat the forceful action multiple times despite immediate tactile and proprioceptive feedback demonstrating spatial impossibility—is one of the most remarkable biomechanical hallmarks of the scale error phenomenon.
6.2 Affective and Expressive Correlates
Equally critical to the identification of scale errors are the child’s concurrent affective and expressive behaviors. A long-standing empirical debate in developmental psychology centered on whether these bizarre behaviors could simply be interpreted as early manifestations of non-verbal pretend play or intentional toddler comedy. However, micro-analytic coding of facial action units, vocal acoustics, and gaze dynamics reveals an affective profile that is entirely antithetical to pretense or humor.
During the execution of a scale error, toddlers consistently display an expression of serious, neutral, or deeply focused concentration. Facial action coding reveals no activation of the zygomatic major muscle (which produces smiling) and no evidence of the Duchenne smile associated with genuine amusement or playfulness. Toddlers do not giggle, chuckle, or produce theatrical vocalizations. There is an utter absence of the exaggerated, playful body postures that children naturally exhibit when engaging in make-believe scenarios.
Furthermore, immediately following the motor failure—when the child’s body fails to enter the car or slips off the miniature slide—their facial affect typically shifts to an expression of profound perplexity, confusion, or mild frustration. The child will frequently freeze in place, staring downward at the miniature object with furrowed brows. A highly characteristic behavioral marker is the subsequent gaze shift: the toddler will abruptly turn their head to make direct eye contact with their parent or the observing experimenter. This look is not the conspiratorial, smiling glance of a child sharing a private joke, but an inquisitive, baffled look that non-verbally communicates: “Why is this not working?” Only after several failed, earnest attempts do older toddlers occasionally transition into a playful demeanor, laughing off the failure as they finally recognize the absurdity of the interaction.
6.3 Frequency and Distribution Across Cohorts
The empirical data gathered across DeLoache’s initial trials and subsequent replication studies provide clear insights into the quantitative frequency and statistical distribution of scale errors across the toddler population. In standard laboratory paradigms utilizing the classic triad of stimuli (chair, car, slide), approximately 50% of typically developing toddlers between the ages of 18 and 30 months commit at least one demonstrable, unambiguously coded scale error during a single brief observation session. When laboratory observation periods are extended, or when children are observed longitudinally in naturalistic home settings, the prevalence rate climbs significantly higher, indicating that the capacity for scale errors is a near-universal developmental phenomenon rather than an idiosyncratic anomaly confined to a small subset of children.
However, the frequency of errors among individual children displays considerable variance. Within a standard testing cohort, researchers observe a clear distinction between “non-responders” (children who do not commit scale errors during the observation window) and “high-frequency errant responders.” While many errant children commit one or two scale errors across the session, a notable subgroup of toddlers exhibits multiple, recurrent scale errors, sometimes committing five, six, or more errors within a ten-minute span, repeatedly attempting to sit on the chair, then moving to the slide, then attempting to insert their foot into the car, and returning to re-attempt the chair.
Contextual variables also modulate occurrence rates. Naturalistic, parent-reported home diaries routinely document scale errors occurring in the child’s everyday environment—such as toddlers attempting to step into a miniature plastic teacup, attempting to climb into a doll’s stroller, or trying to rest their head on a picture of a pillow in a flat storybook. However, the occurrence rate is quantitatively maximized in the laboratory setting. The abrupt, surreptitious substitution of an identical miniature replica immediately following vigorous, rewarding physical interaction with a full-scale referent creates optimal cognitive conditions for an inhibitory breakdown, maximizing the probability that the prepotent motor plan will bypass executive control.
7. Comparative Developmental Trajectories and Age Delimitations
7.1 The Peak Occurrence Window: 18 to 24 Months
The scale error phenomenon exhibits a remarkably sharp, inverted U-shaped developmental trajectory, with its absolute zenith concentrated squarely between 18 and 24 months of age. To understand why this specific six-month window represents the peak vulnerability zone, one must examine the convergence of multiple developmental milestones that occur simultaneously during this stage of early childhood.
First, this period is characterized by the consolidation of locomotor autonomy. By 18 to 20 months, toddlers have moved past the initial, highly precarious stages of learning to walk; their motor programs for locomotion, stair climbing, sitting, and object manipulation have become automated, fluid, and robust. These motor action schemas are no longer tentative, exploratory movements; they are deeply ingrained motor templates stored in procedural memory, ready to be deployed rapidly upon environmental demand.
Second, the 18 to 24-month window coincides with the famous vocabulary explosion and a massive acceleration in semantic object categorization. Children are rapidly learning to group diverse physical objects into generalized functional categories. An armchair, an office chair, a beanbag, and a stool are all integrated under the broad conceptual umbrella of “seating surfaces.” This conceptual generalization is a vital cognitive achievement, but it comes with a transient developmental vulnerability: the child’s ability to recognize the abstract categorical identity of an object temporarily outpaces their executive ability to integrate real-time metric constraints into their motor decisions. The child’s brain is primed to see categorical equivalence everywhere, while their frontal inhibitory and parieto-motor integration circuits are still profoundly immature, creating a developmental “sweet spot” for scale errors to manifest with maximum frequency.
7.2 The Attenuation Phase: 25 to 36 Months
Beyond 24 months of age, the incidence of scale errors drops precipitously. By the time children reach 30 to 36 months, genuine scale errors become vanishingly rare under standard environmental and laboratory conditions. This rapid attenuation reflects the synchronized maturation of several key neurocognitive systems working in concert to stabilize the action-perception loop.
Neurologically, the period between two and three years of age is marked by explosive synaptogenesis and accelerated white matter myelination within the prefrontal cortex, specifically within the circuits connecting the dorsolateral prefrontal cortex, the anterior cingulate cortex, and the basal ganglia. These structural developments endow the 30-month-old child with significantly enhanced top-down inhibitory control. When the ventral stream identifies a miniature replica and primes a functional action schema, the maturing frontal cortex is now capable of rapidly mounting an inhibitory veto, successfully arresting the motor output before it manifests in physical execution.
Simultaneously, the dorsal stream’s online metric updating mechanisms achieve greater computational speed and functional integration with temporal associative cortices. The older toddler’s posterior parietal cortex can instantaneously compute that the miniature replica’s spatial coordinates and physical volume are grossly incompatible with their own bodily dimensions, overriding the ventral categorical impulse almost instantaneously. Furthermore, by 30 to 36 months, children undergo a cognitive revolution in symbolic and pretend play competence. The miniature object is no longer viewed merely as an affordance-bearing physical entity for personal bodily interaction; instead, it is mentally re-categorized as a symbolic prop. The older toddler smoothly redirects their motor actions: rather than attempting to sit in the dollhouse chair themselves, they happily retrieve a doll or a teddy bear and place the doll onto the chair, demonstrating that symbolic mastery has permanently supplanted the sensorimotor scale error.
7.3 Cross-Age Comparisons: Infants Versus Older Preschoolers
To fully appreciate the boundaries of this developmental trajectory, it is instructive to compare the reactions of toddlers to those of younger infants (under 12 months) and older preschool children (aged 3 to 5 years) when presented with the identical miniature stimuli.
Infants under 12 months of age do not exhibit scale errors, but for reasons fundamentally different from preschoolers. A nine-month-old infant lacks the complex, highly consolidated, whole-body motor action schemas required to commit a scale error. An infant of this age does not possess an automated, culturally learned “chair-sitting” schema or an intentional “vehicle-entry” motor program. When an infant encounters a miniature chair or a miniature car, they interact with it using domain-general exploratory sensorimotor schemas appropriate to their developmental stage: they reach for it, grasp it with their fingers, bang it against the floor, transfer it from hand to hand, or bring it to their mouth for oral exploration. The infant fails to commit a scale error because they lack the categorical knowledge and automated motor plans that drive the error in the first place.
At the opposite end of the spectrum, older preschoolers (aged 3 to 5 years) react to miniature replicas with overt metacognitive and metalinguistic awareness. When a four-year-old child is presented with a miniature slide or a miniature armchair in the laboratory setting, they do not attempt to interact with it physically, nor do they look perplexed. Instead, their immediate reaction is almost universally accompanied by humorous affect and explicit verbal commentary. The child will laugh, point at the object, and make explicit declarative statements regarding the scale discrepancy: “Look, it’s tiny!” or “That’s for a baby doll, I’m too big for that!” or “I can’t sit on that, I’d smash it!” For the preschooler, the metric discrepancy is not merely registered by the dorsal visual stream; it is explicitly accessible to conscious, reflective, verbal cognition.
In atypical development, the trajectory of scale errors provides intriguing diagnostic insights. Preliminary investigations in children diagnosed with Autism Spectrum Disorder (ASD) or Developmental Coordination Disorder (DCD) suggest altered scale error patterns. Children with ASD, who frequently exhibit delays in symbolic play alongside atypical local-versus-global visual processing styles, may continue to commit scale errors at chronologically older ages, persisting in treating miniature replicas as physical targets rather than symbolic play items. Conversely, children with severe motor coordination impairments may exhibit a different profile, wherein general motor clumsiness and planning deficits obscure the specific perceptual-action dissociations characteristic of pure scale errors. These cross-population comparisons underscore that the classic scale error is an index of a very specific, normal neurodevelopmental transition.
8. Alternative Interpretations: Pretend Play vs. Genuine Scale Errors
8.1 The Pretense Critique and Counter-Arguments
Following the publication of DeLoache’s 2004 findings, the scale error paradigm faced intense scholarly scrutiny from developmental theorists who questioned whether these behaviors truly represented an unconscious neurocognitive failure. The most prominent and persistent counter-hypothesis was the Pretense Critique. Proponents of this view argued that what DeLoache and her colleagues observed was not a catastrophic visuomotor integration failure, but simply an early, non-verbal form of toddler pretend play. Under this interpretation, when a 20-month-old toddler attempts to sit on a tiny dollhouse chair, they are fully aware of its miniature size and are merely engaging in an imaginative, whimsical game of “make-believe” sitting.
DeLoache and her colleagues mounted a rigorous, multifaceted empirical defense against the pretense critique, centering primarily on micro-behavioral and kinematic analyses. The fundamental argument against the pretense hypothesis lies in the profound biomechanical and affective divergence between genuine pretend play and scale error execution. When toddlers engage in symbolic pretend play—such as pretending to drink from an empty plastic cup, or pretending to sleep by laying their head on a wooden block—their actions are universally characterized by what developmental psychologists term theatricality and motor attenuation. In pretend play, actions are performed daintily, tentatively, and with exaggerated, stylized motions. A toddler pretending to drink does not tilt the cup so violently that plastic jams into their gums; a toddler pretending to sleep does not drop their skull onto the floor with concussive force.
In stark contrast, scale errors involve full biomechanical commitment. As established by force-plate analyses and video coding, toddlers executing scale errors apply genuine, unattenuated physical force. They drop their full body weight onto miniature chairs; they press down with their leg muscles until their joints lock; they attempt to wedge their feet through microscopic openings with aggressive, repetitive force. Furthermore, pretend play is almost universally accompanied by social-communicative cues: toddlers smile, laugh, look up at their parents to verify that the joke is understood, and produce vocalizations indicative of play. Scale errors, as demonstrated repeatedly, occur in the total absence of play affect; the children exhibit serious, focused, and ultimately perplexed expressions. The pretense critique simply fails to explain why a child would intentionally make-believe an action to the point of repeatedly falling over, losing balance, or exhibiting genuine distress upon physical failure.
8.2 Slippage and Pragmatic Communication Theories
A second major alternative framework proposed to explain scale errors centers on theories of conceptual slippage and pragmatic communication. Advanced by researchers interested in socio-communicative development, this perspective suggests that scale errors are an artifact of the novel, somewhat artificial communicative demands of the laboratory environment. According to this hypothesis, young children placed in an experimental room with unfamiliar adults experience an implicit social pressure to act. When the full-sized, highly engaging toys are suddenly replaced with miniature versions, the child experiences a pragmatic dilemma: they want to comply with the implicit demand to “play,” but the room offers only inadequate objects.
Under this social compliance model, the toddler’s behavior represents a form of “conceptual slippage.” The child recognizes that the miniature car is not an appropriate vehicle, but because the experimental context demands action, the child allows their behavioral boundaries to slip, enacting the most proximate functional schema available in an effort to satisfy perceived experimental demand characteristics. The child is not suffering from a visual-motor dissociation; rather, they are attempting a pragmatic, best-effort behavioral response in an ambiguous social context.
To dismantle this pragmatic communication hypothesis, DeLoache and independent replication teams conducted elegant experimental variations designed to control for adult presence, social cuing, and demand characteristics. In these controlled paradigms, toddlers were observed through one-way mirrors while completely alone in the testing room, or in naturalistic home environments where no unfamiliar adults were present and no experimental demands were imposed. Scale errors continued to occur with comparable frequency and intensity. Even when toddlers believed they were completely unobserved, they earnestly attempted to sit on dollhouse chairs and wedge themselves into miniature cars. These findings decisively demonstrated that scale errors are not communicative strategies designed to appease adult observers, but internally generated sensorimotor actions arising directly from the child’s autonomous cognitive processing.
8.3 Perceptual Processing Deficits vs. Action Control Deficits
A third theoretical challenge posited that scale errors might simply stem from an elementary perceptual processing deficit. Could it be that toddlers between 18 and 24 months of age simply possess immature visual acuity, poor depth perception, or an inability to accurately discriminate physical size at a distance? If a child cannot visually distinguish a 10-centimeter chair from a 50-centimeter chair, then their attempt to sit on the miniature replica would represent a primary sensory error rather than an interesting perception-action dissociation.
This perceptual deficit hypothesis was swiftly and definitively refuted by DeLoache and her team through a series of ingenious perceptual choice control trials. Immediately following an experimental trial in which a toddler committed a scale error on a miniature object, the experimenter placed the miniature replica side-by-side on the floor with its full-sized, functional counterpart. The experimenter then issued simple, neutral verbal prompts, such as: “Can you bring me the big chair?” or “Sit in the big chair,” or conversely, “Bring me the little car.”
The results of these choice tests were definitive and unequivocal: toddlers who had just committed a scale error on the miniature object exhibited a 100% success rate on the perceptual size discrimination task. They instantly pointed to, retrieved, or sat upon the requested object based on size terminology. They demonstrated flawless visual acuity and precise, conscious size discrimination. They knew without hesitation which object was large and which was small. These findings established the foundational paradox of the scale error: the toddler’s visual system possesses perfectly intact size perception when probed via conscious, reflective choice tasks, yet that very same size information completely fails to integrate into their motor planning system during rapid, unreflective action execution. It is an action control deficit, not a perceptual deficit.
9. Replications, Variations, and Cross-Cultural Perspectives
9.1 Independent Laboratory Replications
The scientific robustness of any landmark psychological discovery depends on its capacity to withstand independent replication across disparate laboratories and research methodologies. Following DeLoache’s 2004 publication, the scale error paradigm was subjected to extensive replication efforts across multiple international developmental laboratories in North America, Western Europe, and East Asia (e.g., Ware et al., 2006; Rosengren et al., 2009; Casler et al., 2011). These independent investigations overwhelmingly corroborated DeLoache’s core empirical findings, confirming that scale errors represent a reliable, statistically replicable developmental phenomenon.
These independent replications consistently replicated the primary effect sizes, confirming that roughly 40% to 55% of toddlers in the 18 to 30-month age bracket exhibit demonstrable scale errors in laboratory settings, with the peak incidence consistently localized around 20 to 24 months. Furthermore, these replications established the structural invariance of the behavioral typologies: across cultures and laboratories, sitting on miniature chairs and attempting to enter miniature vehicles remained the most frequent and easily elicited error typologies.
However, these replication efforts also identified critical environmental and procedural variables that modulate scale error frequency. Researchers discovered that the duration of the distraction interval played a vital role: if the time delay between interacting with the large object and encountering the miniature replica exceeded five minutes, the incidence of scale errors declined, suggesting that the prepotent motor priming induced by the large object undergoes temporal decay. Conversely, increasing the physical distance between the child and the miniature object upon re-entry slightly reduced error rates, as the extended approach time provided a broader temporal window for dorsal-stream metric processing to intervene and arrest the motor cascade.
9.2 Variations in Object Familiarity and Context
Subsequent iterations of the paradigm expanded beyond the classic triad of chairs, slides, and cars to investigate how varying degrees of object familiarity, material composition, and environmental context influence scale error rates. Studies introducing highly novel, unfamiliar artifacts—objects specifically invented by researchers with arbitrary shapes and newly taught functional affordances—revealed that scale errors are significantly less frequent on novel items compared to ubiquitous domestic objects. A prepotent action plan requires extensive, habitual sensorimotor reinforcement; because toddlers possess hundreds of hours of physical experience with chairs and shoes, those motor templates are primed with vastly greater electrophysiological potency than newly acquired actions on novel objects.
Other variations systematically manipulated the physical cues of the miniature stimuli, altering their color, material texture, and weight authenticity. When researchers presented miniature chairs made of clear, transparent acrylic plastic, or chairs painted in colors radically divergent from the full-sized referents, scale error rates dropped markedly. The reduction in low-level visual feature overlap disrupted the ventral stream’s rapid, bottom-up categorization, preventing the instantaneous activation of the prepotent sitting schema and allowing the child’s reflective attention to intervene.
In contemporary developmental research, the scale error paradigm has expanded into the digital realm, examining toddler interactions with touchscreens, tablets, and virtual interfaces. In these modern variations, toddlers are observed interacting with photorealistic, two-dimensional digital representations of objects displayed on digital screens. Researchers have documented intriguing digital analogues of scale errors: toddlers frequently attempt to physically grasp a 2D digital image of a ball off the screen surface, or attempt to place their foot onto a digital depiction of a step displayed on an interactive floor tablet. These touchscreen scale errors demonstrate that the perceptual-action dissociation identified by DeLoache is not merely an artifact of plastic toys, but an inherent vulnerability of the developing brain whenever compelling pictorial or miniature cues collide with immature motor control networks.
9.3 Cross-Cultural Generalizability
A perennial critique of modern developmental psychology is its heavy reliance on so-called WEIRD demographics (Western, Educated, Industrialized, Rich, and Democratic societies). Because children in WEIRD cultures are uniquely submerged in consumer environments saturated with manufactured toys, miniature dollhouses, and scaled-down replica playthings, critics questioned whether scale errors were simply an artifact of Western toy socialization. Do children raised in non-Western, rural, or low-resource settings—where commercial, geometrically precise miniature replica toys are virtually non-existent—display scale errors?
To resolve this fundamental question, cross-cultural developmental researchers conducted comparative field studies in diverse cultural settings, including rural indigenous communities in South America, traditional agrarian villages in sub-Saharan Africa, and non-industrialized communities in India. The findings provided striking evidence for the universal neurobiological basis of the scale error. Even in environments devoid of commercial toy replicas, children between 18 and 24 months of age routinely commit scale errors when presented with miniature objects or when encountering natural miniature entities in their daily environment.
In rural subsistence settings, toddlers were observed attempting to wash their hands in tiny bottle caps, attempting to put their feet into tiny discarded sandals worn by infant siblings, or attempting to climb into tiny woven storage baskets intended for grain. While the specific cultural artifacts differed, the cognitive architecture of the error remained entirely invariant: the categorical identification of an object triggered an automated motor program that failed to integrate real-time metric scale. These cross-cultural findings decisively demonstrated that the scale error is not a cultural byproduct of Western commercial toy markets, but a universal developmental milestone reflecting the cross-species architecture of the developing primate visual system.
10. Theoretical Implications for Representational Insight and Dual Representation
10.1 Integration with DeLoache’s Dual Representation Theory
The empirical confirmation of the scale error paradigm provided the vital missing link in Judy DeLoache’s overarching theoretical architecture, serving as the profound conceptual inverse of her celebrated Dual Representation Theory. To understand this theoretical synthesis, one must compare the cognitive demands of the classic “model room” task with the cognitive demands of the scale error experiment.
In the model room task, the toddler’s primary cognitive failure stems from the inability to see the object as a symbol. The miniature model room is so physically captivating, tangible, and real that the 30-month-old child cannot look past its physical reality to mentally represent its distal, symbolic meaning. The physical object “eclipses” the symbol. The child is trapped by the concrete objectness of the artifact.
In the scale error paradigm, an exact reciprocal breakdown occurs: the toddler exhibits the inability to see the symbol purely as a physical object with its own distinct metric boundaries. When confronted with the miniature replica armchair, the toddler’s brain is so utterly overwhelmed by the abstract, categorical representation of “chair-ness” (a symbol of function) that they completely ignore the concrete, physical metrics of the specific object resting on the floor. In the model room, the physical object obscures the symbol; in the scale error, the mental representation of the category completely obliterates the physical reality of the object! Both phenomena represent the twin poles of an immature representational system struggling to negotiate the delicate, dialectical boundary between abstract mental categories and concrete physical objects in the material world.
10.2 Insights into Action Representation and Affordance Perception
DeLoache’s scale error discoveries forced a fundamental re-evaluation of ecological psychology’s concept of affordance perception. In classical Gibsonian ecological theory, affordances were conceptualized as directly perceived, immutable relations between the organism’s bodily dimensions and the environmental layout. Gibsonians maintained that an organism does not perceive abstract geometry and then calculate whether an action is possible; rather, the action possibility itself is the primary perceptual reality. A surface is directly perceived as “walk-on-able” or “sit-on-able.”
The scale error delivers an insurmountable empirical challenge to pure Gibsonian direct perception. If affordances were directly and flawlessly perceived through ecological visual invariants, a toddler could never, under any circumstances, perceive a 10-centimeter dollhouse chair as affording “sitting-on” for their 12-kilogram body. The very occurrence of a scale error proves that affordance perception in developing humans is not an unmediated, infallible ecological pickup of environmental information. Instead, affordance perception is a computationally complex, mediated process that requires the seamless, real-time integration of two completely distinct cognitive representations: knowing what an object is for (semantic, categorical affordance) versus knowing how to act on it right now (metric, biomechanical affordance).
Scale errors expose the profound mental dissociation between these two systems. They demonstrate that affordance representation is hierarchically structured and embodied. In the young toddler, the “what-affordance” (semantic knowledge that chairs are for sitting) can detach from the “how-affordance” (the physical calculation of body-to-object scale). This profound insight has enriched contemporary theories of embodied cognition, demonstrating that our sensorimotor grounding in the physical world is not an innate, pre-packaged computational module, but an actively constructed, neurologically vulnerable developmental achievement.
10.3 The Modularity of the Developing Mind
In the broader landscape of cognitive science, the scale error paradigm serves as one of the most powerful empirical broadsides against traditional, domain-general models of human cognitive development. For decades, classical cognitive theories—from Piagetian constructivism to broad structural developmentalism—posited that the mind matures in broad, synchronized cognitive stages. Under a domain-general model, an individual child operates at an overarching cognitive level; a child who understands object permanence or categorical identity was assumed to apply that cognitive stage uniformly across perception, reasoning, and motor action.
The scale error provides incontrovertible behavioral evidence in favor of developmental modularity and asynchronous cognitive development. It demonstrates that the mind does not mature as a unified, monolithic entity. Instead, the developing brain is composed of semi-autonomous, modular cognitive and neurological sub-systems—specifically, the bifurcated visual pathways, the motor execution cascades, the semantic categorization networks, and the prefrontal inhibitory circuits—that develop at dramatically different chronological rates.
Seen through the framework of Dynamic Systems Theory, advanced by developmentalists such as Esther Thelen and Linda Smith, the scale error is not a hardwired “program” or a static cognitive deficit. Rather, it is an emergent developmental phenomenon. It arises dynamically at the behavioral crossroads where rapid locomotor independence, explosive semantic categorization, immature prefrontal inhibitory control, and unmyelinated dorsal-ventral connectivity briefly intersect in time. When these asynchronous systems collide in a specific physical environment, the scale error temporarily emerges as an inevitable property of a cognitive system undergoing radical self-organization.
11. Methodological Critiques, Confounds, and Experimental Limitations
11.1 Laboratory Ecological Validity Concerns
Despite its historic impact, the classic scale error paradigm has been subject to rigorous methodological critiques regarding ecological validity, observational confounds, and experimental design constraints. Foremost among these is the concern regarding laboratory ecological validity. In DeLoache’s standardized 2004 protocol, the experimental environment is an undeniably artificial, highly staged space. The rapid, surreptitious substitution of full-sized objects with identical miniature replicas immediately following an intense familiarization phase creates an environmental setup that children rarely encounter in ordinary domestic life.
Critics point out that this specific sequence of events introduces a massive novelty and priming confound. By actively encouraging the child to slide, sit, and drive during the familiarization phase, the experimenters intentionally super-charge the motor cortex with prepotent activation, and by subsequently removing the child and presenting identical miniature replicas, they maximize the probability of an inhibitory failure. While this artificiality is a common feature of rigorous experimental control, skeptics argue that it may inflate the apparent frequency and severity of scale errors relative to their true baseline occurrence in unmanipulated, naturalistic environments.
Furthermore, the physical presence of caregivers in the testing room introduces potential unconscious behavioral cuing (the classic Clever Hans effect). Even when instructed to remain completely neutral, parents sitting in the corner of a laboratory room often exhibit subtle, non-verbal postural shifts, micro-facial expressions of anticipation, or altered gaze directions when their child approaches a miniature object. These subtle socio-communicative cues could inadvertently scaffold, encourage, or prolong the child’s bizarre motor interactions, confounding the interpretation of the behavior as an entirely autonomous cognitive error.
11.2 Coding Subjectivity and Ambiguity in Motor Intent
A second major methodological challenge involves the subjectivity of behavioral coding and the inherent ambiguity of inferring internal mental states solely from observable motor trajectories. A central pillar of the scale error definition is that the child must possess “serious motor intent.” However, discerning the precise boundary between an earnest, serious motor attempt and an exploratory, sensory-tactile physical probe is extraordinarily difficult when analyzing the fluid, idiosyncratic movements of a 20-month-old toddler.
For example, if a toddler bends their knees over a miniature chair, hesitates, and brushes their buttocks against the seat before standing back up, does this constitute an earnest scale error halted by rapid dorsal feedback, or does it represent an intentional, playful physical touch? While DeLoache established stringent coding schemas with high inter-rater reliability metrics ($kappa > 0.90$), critics argue that human coders—primed by the dramatic hypothesis of the study—may naturally suffer from confirmation bias, unconsciously over-interpreting ambiguous exploratory postures as full-blown scale errors.
To overcome this limitation, contemporary cognitive scientists have demanded the integration of objective, continuous biomechanical tracking technologies. Modern replication studies have begun incorporating high-speed multi-camera motion capture systems, wearable kinematic inertial measurement units (IMUs), and floor-embedded piezoelectric force plates. These advanced technologies eliminate human coding subjectivity by quantifying exact kinetic parameters: the precise center-of-pressure shifts, ground reaction forces, and joint torque dynamics during the approach and contact phases. While early force-plate studies have largely corroborated DeLoache’s qualitative coding, the reliance on purely visual video coding in earlier literature remains an acknowledged methodological limitation.
11.3 Sample Size and Selection Biases in Developmental Psychology
Finally, the scale error literature—like much of developmental cognitive psychology—faces methodological constraints regarding sample size, statistical power, and demographic selection biases. DeLoache’s foundational 2004 study was based on a total cohort of 54 toddlers, divided across multiple experimental age groups and stimulus conditions. When small cohorts are stratified across cross-sectional age bands, the statistical cell sizes for specific error typologies (such as sliding errors versus car-entry errors) become relatively modest.
Furthermore, the participants in these initial laboratory cohorts were drawn almost exclusively from university-adjacent communities in middle-to-upper-class suburban settings—a quintessential WEIRD demographic. Families who have the socio-economic freedom, flexible time, and transportation resources to volunteer for infant laboratory testing represent a highly privileged, non-representative slice of the global human population. These children routinely inhabit domestic environments filled with expensive educational toys, books, and adult-scaffolded play interactions, factors that may subtly influence symbolic comprehension and spatial cognitive development.
From an epidemiological and statistical perspective, the individual base rate of scale errors also presents analytical challenges. While roughly 50% of children commit an error during a testing session, the total number of discrete scale error events recorded per child is often quite low (averaging one to two errors per session). Conducting complex parametric statistical analyses and modeling individual developmental variations on behavioral phenomena with low individual base rates introduces statistical noise and heightens the risk of Type I or Type II errors. Modern developmental neuroscience recognizes that establishing definitive structural-behavioral links requires large-scale, multi-site collaborative cohorts (such as the “ManyBabies” consortium models) to ensure robust statistical power.
12. Contemporary Legacy and Future Directions in Cognitive Developmental Neuroscience
12.1 Neuroimaging Advances and Empirical Corroboration
In the two decades since DeLoache’s seminal paper, the scale error paradigm has evolved from a purely behavioral curiosity into a vital experimental model for modern cognitive developmental neuroscience. Historically, verifying the neural mechanisms underlying scale errors was severely constrained by the impossibility of placing active, moving toddlers inside traditional, motion-sensitive functional Magnetic Resonance Imaging (fMRI) scanners. Today, revolutionary non-invasive neuroimaging technologies have bridged this empirical chasm.
Foremost among these neuroimaging breakthroughs is the application of functional Near-Infrared Spectroscopy (fNIRS) to mobile toddler cohorts. fNIRS utilizes lightweight, head-mounted optical sensor caps that measure localized cortical hemodynamic responses (changes in oxygenated and deoxygenated hemoglobin) while permitting the child to move freely, walk, and manipulate objects in an ecological play space. Recent fNIRS studies investigating toddler action planning have provided direct empirical corroboration for Milner and Goodale’s dual-stream model: when toddlers approach miniature replica objects, optical probes positioned over the inferotemporal cortex reveal robust, immediate hemodynamic activation (reflecting intact categorical identification), while probes positioned over the dorsolateral prefrontal cortex and posterior parietal regions show delayed, attenuated, or disorganized activation patterns in children who proceed to commit scale errors.
Simultaneously, advancements in pediatric structural MRI and diffusion tensor imaging (DTI) have enabled neuroscientists to track the physical myelination of white matter pathways in longitudinally tracked toddler cohorts. Structural MRI studies reveal that the chronological cessation of scale errors—the moment in development when a child permanently stops attempting to sit on miniature chairs—correlates precisely with the structural maturation and fractional anisotropy (FA) thresholds of the superior longitudinal fasciculus and the corpus callosum. These white matter tracts facilitate the high-speed, inter-hemispheric, and anterior-posterior cortical communication required for the prefrontal cortex to exert top-down inhibitory control over motor execution. The scale error has thus transitioned from an observational behavioral insight into an empirically mapped neurobiological milestone.
12.2 Robotics, Artificial Intelligence, and Computer Vision Analogues
The contemporary legacy of Judy DeLoache’s scale error experiment extends far beyond developmental psychology, exerting a profound and unexpected influence on the fields of robotics, artificial intelligence, and computer vision. As roboticists strive to engineer autonomous humanoid robots capable of navigating human environments and manipulating physical tools, they have repeatedly encountered a computational architecture problem that directly mirrors the toddler’s scale error: the computational dissociation between semantic object recognition and embodied physical affordance calculation.
In classical computer vision systems powered by deep convolutional neural networks (CNNs), an AI model can be trained to recognize an object with astonishing accuracy. A computer vision algorithm can effortlessly identify an armchair in an image, tagging it as “chair” with 99% confidence. However, when an autonomous embodied robotic agent equipped with this categorical vision network is deployed in a physical room containing a miniature toy chair, the robot will frequently attempt to position its heavy mechanical chassis over the miniature chair to “sit,” resulting in mechanical tipping, structural collision, or motor failure. The robot commits an algorithmic scale error! The deep neural network identifies the categorical semantic label (“what”), but the robot’s real-time kinematic path planner fails to compute the metric, load-bearing affordance (“how”) relative to its own physical hardware.
To overcome this fundamental failure mode, contemporary roboticists have directly adopted DeLoache’s insights to design bi-stream neural architectures for embodied AI. Modern robotic systems are intentionally partitioned into dual, cross-communicating computational pipelines: an allocentric semantic classification network that identifies object classes, and an egocentric geometric point-cloud network that computes absolute spatial volumes, surface normals, and mechanical affordances in real time. DeLoache’s developmental psychology paradigm has provided computer scientists with an evolutionary roadmap: by understanding how the human toddler’s brain gradually bridges the divide between ventral categorization and dorsal motor control, engineers are learning how to build adaptive, error-free visuomotor algorithms for next-generation autonomous robotics.
12.3 Pedagogical, Clinical, and Practical Applications
Finally, the scale error paradigm has generated enduring, transformative applications across early childhood pedagogy, toy manufacturing design, clinical screening, and parenting education. In the realm of early childhood learning environments, DeLoache’s discoveries revolutionized our understanding of how toddlers interact with physical learning materials. Historically, preschool and daycare environments frequently mixed full-scale functional furniture with miniature decorative models, creating continuous cognitive friction for children undergoing the peak scale error vulnerability window. Modern pedagogical architecture emphasizes spatial clarity, providing environments where objects are clearly segregated into functional tools scaled to the child’s body versus designated, highly stylized symbolic play spaces, minimizing motoric confusion and optimizing executive learning.
In the commercial domain of toy design and child safety engineering, the scale error paradigm established critical safety imperatives. Prior to DeLoache’s work, toy manufacturers often designed miniature toy vehicles and doll furniture with hyper-realistic materials, rigid edges, and structural components that appeared identical to full-scale furniture. Because developmental science proved that toddlers between 18 and 24 months will earnestly apply their full body mass to these objects with ballistic force, international safety regulatory bodies updated engineering standards. Miniature toys designed for the toddler demographic must now undergo rigorous structural load testing and impact resistance standards to ensure that when a 20-month-old toddler inevitably drops their full body weight onto a miniature plastic chair or presses their foot into a miniature toy car, the artifact will not shatter into sharp, hazardous plastic shards.
Clinically, the scale error profiling paradigm is emerging as a potential non-invasive behavioral biomarker in early neurodevelopmental screening. Because the emergence and subsequent attenuation of scale errors are governed by the precise, predictable structural maturation of prefrontal inhibitory pathways and parieto-temporal white matter tracts, deviations from the standard developmental trajectory can provide early diagnostic clues. A child who continues to exhibit frequent, persistent, high-force scale errors past 36 or 42 months of age—in the complete absence of symbolic pretense or metalinguistic humor—may be exhibiting early behavioral markers of executive dysfunction, inhibitory control delay, or atypical visuomotor integration, warranting comprehensive neurodevelopmental evaluation for conditions such as ADHD, Developmental Coordination Disorder, or Autism Spectrum Disorder.
Ultimately, Judy DeLoache’s scale error experiment endures as one of the most brilliant, transformative paradigms in the history of developmental science. It took a fleeting, bizarre, and often humorous childhood behavior—a toddler earnestly trying to squeeze their body into a microscopic toy car—and extracted from it a profound, elegant, and revolutionary window into the architecture of the human mind. The scale error stands as a timeless testament to the fact that growing up is not simply a matter of learning more things about the world; it is the extraordinary, delicate, and wondrous biological journey of weaving together the disparate threads of perception, action, and thought into a unified human consciousness.
Conclusion
The scale error phenomenon, as systematically isolated and articulated by Judy DeLoache and her colleagues, represents one of the most illuminating chapters in modern developmental psychology and cognitive neuroscience. Far from being a trivial behavioral aberration or an amusing instance of childhood foolishness, the scale error exposes the deep, intricate scaffolding that underpins the human mind. It lays bare the remarkable reality that our perception of the world is not a single, seamless, uniform stream of conscious experience, but rather the hard-won synthesis of parallel, asynchronously developing neural networks that must learn to communicate across developmental time.
Through rigorous empirical methodology, DeLoache demonstrated that between the ages of 18 and 30 months, human toddlers inhabit a unique neurodevelopmental landscape. Their ventral visual pathways and semantic categorization systems have advanced to the point of rapid, sophisticated conceptual generalization—they instantly recognize the universal functional affordances of chairs, cars, and slides. Yet, their dorsal visuomotor pathways and prefrontal inhibitory networks are still undergoing the slow, structural processes of synaptogenesis and axonal myelination. In this developmental interregnum, the child’s brain is capable of knowing what an object is for while remaining temporarily powerless to stop an inappropriate motor command from firing toward an impossible target.
The legacy of DeLoache’s 2004 Science experiment continues to reverberate across disciplinary boundaries. It provided the crucial complementary pillar to the Dual Representation Theory, showing that children can be blinded by abstract categories just as easily as they can be blinded by concrete objects. It forced ecological psychology to reconsider the complex, mediated nature of affordance perception, and it delivered decisive evidence against monolithic, domain-general models of cognitive development. Today, as neuroscientists map these pathways with mobile neuroimaging and roboticists construct dual-stream vision systems for artificial intelligence, the image of a 20-month-old toddler earnestly trying to sit on a two-inch dollhouse chair remains an enduring, poignant symbol of the developing human brain: an extraordinary cognitive architecture actively striving, stumbling, and organizing itself on the wondrous journey toward maturity.
References
Casler, K., Terziyan, T., & Greene, K. (2011). Toddlers’ scale errors: A direct test of the dual-stream hypothesis. Journal of Experimental Child Psychology, 108(4), 812–820. https://doi.org/10.1016/j.jecp.2010.10.003
DeLoache, J. S. (1987). Rapid change in the symbolic functioning of very young children. Science, 238(4833), 1556–1557. https://doi.org/10.1126/science.2446392
DeLoache, J. S. (2000). Dual representation and young children’s use of scale models. Child Development, 71(2), 329–338. https://doi.org/10.1111/1467-8624.00148
DeLoache, J. S., Miller, K. F., & Rosengren, K. S. (1997). The credible shrinking room: Very young children’s performance with symbolic and nonsymbolic relations. Psychological Science, 8(4), 308–313. https://doi.org/10.1111/j.1467-9280.1997.tb00443.x
DeLoache, J. S., Uttal, D. H., & Rosengren, K. S. (2004). Scale errors offer evidence for a perception-action dissociation early in life. Science, 304(5673), 1027–1029. https://doi.org/10.1126/science.1093567
Diamond, A. (1990). The development and neural bases of higher cognitive functions. Annals of the New York Academy of Sciences, 608(1), 267–317. https://doi.org/10.1111/j.1749-6632.1990.tb48898.x
Gibson, E. J., & Walk, R. D. (1960). The “visual cliff”. Scientific American, 202(4), 64–71. https://doi.org/10.1038/scientificamerican0460-64
Gibson, J. J. (1979). The ecological approach to visual perception. Houghton Mifflin.
Goodale, M. A., & Milner, A. D. (1992). Separate visual pathways for perception and action. Trends in Neurosciences, 15(1), 20–25. https://doi.org/10.1016/0166-2236(92)90344-8
Milner, A. D., & Goodale, M. A. (1995). The visual brain in action. Oxford University Press.
Piaget, J. (1952). The origins of intelligence in children. International Universities Press.
Piaget, J. (1954). The construction of reality in the child. Basic Books.
Rosengren, K. S., Carmichael, C., Schein, S. S., Anderson, K. N., & Gutierrez, N. (2009). A hot tub for a doll? Scale errors in toddlerhood. Journal of Cognition and Development, 10(3), 194–212. https://doi.org/10.1080/15248370903155825
Thelen, E., & Smith, L. B. (1994). A dynamic systems approach to the development of cognition and action. MIT Press.
Ungerleider, L. G., & Mishkin, M. (1982). Two cortical visual systems. In D. J. Ingle, M. A. Goodale, & R. J. W. Mansfield (Eds.), Analysis of visual behavior (pp. 549–586). MIT Press.
Uttal, D. H., DeLoache, J. S., & Rosengren, K. S. (2007). Scale errors: What they tell us about the development of action and representation. Current Directions in Psychological Science, 16(5), 243–247. https://doi.org/10.1111/j.1467-8721.2007.00513.x
Ware, E. A., Uttal, D. H., & DeLoache, J. S. (2006). Everyday experience and scale errors. Infancy, 9(2), 223–236. https://doi.org/10.1207/s15327078in0902_7
Zelazo, P. D., & Müller, U. (2002). Executive function in typical and atypical development. In U. Goswami (Ed.), Blackwell handbook of childhood cognitive development (pp. 445–469). Blackwell Publishing.