**1. Introduction**

Relative to goal-directed actions (manifested with behavioral responses), *selective-set* (as defined by Kahneman and Triesman [1]) is a mechanism of selective attention which underlies the ability of detecting task-relevant target information while ignoring (temporally simultaneous or separate) irrelevant information within a sequence of stimuli [1]. In many variations of the selective-set paradigm, speed and accuracy of behavioral response generally improve with age (for review see [2]), and one

consistently replicated finding is a sharp developmental transition observed in children between 3 and 5 years of age—he developmental period usually known as "preschool". This transition is often implied as a critical period for detailing neurofunctional mapping [3,4], and for understanding typical and atypical development of executive attention during the lifespan [5].

The preschool transition has also been reported in studies in which event-related potentials (ERPs) were measured concurrently with different variants of the selective-set paradigm. An important component of selective-set detection is *response inhibition*, which has been studied in the laboratory by using deviant target detection tasks such as the go/no-go task, continuous performance task (CPT) and the stop-signal task [4–6]. Specifically, in ERP studies, predominantly in the auditory modality, larger amplitudes of the N2 with (200–300 ms) component have been found for successful responses to no-go trials compared to go trials, which are similar to those found in adults [6–8]. However, the N2 component of young children is usually observed between 250 and 500 ms after stimulus onset [9,10] and the no-go N2 effect is larger and more widely distributed across the fronto-parietal electrodes [9,11]. Similar findings in both morphology and latency differences have been reported for other visual processing and visual search tasks [12,13] and in a body of converging findings on error monitoring during visual go/no-go tasks related to error positivity (Pe) and negativity (ERN) components [14–16].

To date, Ridderinkhof and van der Stelt [17] conducted the most exhaustive review (including reviews of older studies) of adult and children's ERP data, which shows that signatures such as the N200 and P300-like (especially the P3b component) were reported in experiments using deviant target detection paradigms such as visual oddball tasks. Importantly, these studies included small convenience samples of 5–6 year-old children (sometimes in groups of broader age ranges, i.e., 4–7 year-olds). Late negative Nc (410 ms) and positive Pc (900 ms) waves (see [18]) as well as Slow Waves were observed in the young participants, and further in conditions where targets were novel stimuli. During the preschool and kindergarten years, the timing of the P300 is significantly slower, peaking on average around 700 ms and ranging between about 600 and 900 ms. Both children and young adults show greater P3 amplitudes to target, attended stimuli relative to non-target, unattended stimuli, and their topographical organizations are qualitatively similar in distribution across posterior electrodes [18].

Based on a modeling analysis of the reviewed studies, Ridderinkhof and Van der Stelt [17] concluded that attentional selective-set is essentially "adult-like" in preschool children, but the age differences in ERP wave morphology and latency may indicate that processing speed and efficiency undergo developmental improvement. Thus, their conclusion implies a form of developmental homology, i.e., an equivalence of structure and functions at two different developmental moments [19]—preschool and adulthood—capturing some neurocognitive aspects of selective-set, but not others (such as, for instance, those related to the execution of response). How this correspondence may practically translate to specific neural mechanisms is still an open question.

According to an influential interpretation [20,21], preschool transition involves a shift from involuntary detection and response towards novelty, to voluntary control of attention and of response to target (including withholding response or attenuating distractors' interference, or inhibitory control, in favor of more appropriate target response). The shift to voluntary control is usually attributed to the relatively early development of response inhibition and linked with the functional maturation of the frontal system, which is assumed to reflect key changes in the connectivity of the prefrontal cortex (PFC) during the preschool period [22–24]. Indeed, the preschool period shows a dramatic maturation of axonal density and myelination of structures supporting visuomotor functions in the frontal-striatum and fronto-basal ganglia networks, and in the fast propagating fibers of the callosal connections of the motor corticospinal system [25–34].

Voluntary control of visual selective attention has been most recently defined as "top-down" driven (i.e., "regulated by the working memory central executive") neuronal activity, which is directed to selectively enhance relevant target information and attenuate potential distractors [35]. *Enhancement* is generally associated with larger neuronal and significantly higher electrophysiological activity (see review in [36]) or eye-movemen<sup>t</sup> activity [37] concurrent to targets as compared to distractors.

In the present study, we tested the twofold hypothesis that voluntary selective-set detection in preschool children may be associated specifically with enhancement of neural response to target (enhancement hypothesis). Furthermore, we tested the hypothesis that in preschool children this mechanism can be described by some of the similar features, structural and functional, which are attributed to adults (neurofunctional homology).

In the first and second part of the study, we reanalyzed and modeled ERP data from preschool children on a sustained visual detection task, specifically probing set-selection. We expected to find converging evidence that children's ERP activity concurrent with the target would generally show higher amplitudes, as compared to ERP activity concurrent with the distractor.

In the third part of the study, we extended an ERP-based neurocognitive modeling approach to test the extent to which: 1) The pattern and timing of the preschoolers' actual ERP responses to target and distractor could be explained by a simulated adult model of ERP activity (functional homology); and 2) the dipoles estimated from preschool children's ERP activity could approximate the adult spatiotemporal simulation of estimated ERP generators (structural homology).

For periods of the task examined here, the ERP correlates of the selective-set process were confounded with those of the execution of manual response. Nonetheless, by triangulating peak analysis, source dipole modeling, simulation, and spiking modeling, we sought to partition distinct temporal intervals in which we could discriminate with reasonable degree of probability the neural processes recruited predominantly for target enhancement from those recruited for the planning of release/inhibition of response.

#### **2. Materials and Methods**

#### *2.1. Part 1: Reanalysis of Children's ERP Data*

In the first part of this study, we reanalyzed ERP data collected from preschool children who were tested on an adapted computerized version of Akshomo ff's visual sustained detection task [38]. First, using a binning-averaging technique (vincentization), we tested whether ERP amplitudes for targets could be characterized as enhanced activity compared to the ERP amplitudes for distractors. Data vincentization made sure these di fferences were parametrized; namely, they did not depend on phase or time delays between the two conditions' waveforms, and neither on other individual's distribution variations from the grand average distribution.
