Exploring early-stage orienting behavior using an eye tracker for attention deficit hyperactivity disorder classification

Exploring early-stage orienting behavior using an eye tracker for attention deficit hyperactivity disorder classification | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Exploring early-stage orienting behavior using an eye tracker for attention deficit hyperactivity disorder classification Seonmi Lee, Sangil Lee, Inji Jeong, Jaehyun Jeong, Hyoju Park, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7191610/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 26 Feb, 2026 Read the published version in Scientific Reports → Version 1 posted 12 You are reading this latest preprint version Abstract Exploring early-stage orienting behavior is essential for elucidating the behavioral mechanisms underlying attentional shifts in attention deficit hyperactivity disorder (ADHD). However, traditional tasks lacking eye-tracking data often obscure these mechanisms. This study investigates low-level attentional shifting in ADHD using a simplified gaze-cueing task and explores classification markers via eye movement. Eye-tracking data were collected from 44 typically developing children and 28 children diagnosed with ADHD. We constructed a logistic regression model for classification purposes. Eye movement data alone yielded an accuracy of 0.84, comparable to the accuracy achieved using combined eye-tracking and behavioral data (0.87), underscoring the sensitivity of gaze-based features. Children with ADHD exhibited significantly prolonged fixation (p = .02, d = 0.80) and marginally reduced saccade frequency (p = .06, d = − 0.52) during target detection, indicating delayed attentional shifting and diminished goal-directed attention. Prolonged fixation during target detection behavior emerged as the strongest predictor, correlating with both inattention and hyperactivity (r = .46; r = .36; both p < .01). Additionally, children with ADHD demonstrated lower joint attention and a greater reliance on peripheral vision. These findings highlight distinct gaze patterns under low cognitive load, revealing subtle mechanisms of executive dysfunction and potential early classification markers. Health sciences/Diseases Biological sciences/Neuroscience Biological sciences/Psychology Social science/Psychology early-stage orienting behavior ADHD eye movement attention shifts classification gaze-cueing task Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction Attention deficit hyperactivity disorder (ADHD) is a prototypical neurodevelopmental disorder characterized by persistent symptoms of inattention, hyperactivity, and impulsivity. Although it typically manifests during childhood, these symptoms often persist into adolescence and adulthood [ 1 , 2 ]. ADHD significantly impacts academic performance and daily functioning, primarily due to difficulties in sustaining attention, and is closely associated with various cognitive deficits. Theoretically, deficits in attention and impulsivity associated with ADHD have been attributed to impairments in executive functions [ 3 ]. Executive function refers to higher-order cognitive processes that regulate goal-directed behavior by integrating stimuli and controlling responses. These processes include stimulus prioritization, response inhibition, and cognitive shifting to achieve specific goals [ 4 ]. Barkley (1997) notably suggested that reduced prefrontal cortex activity in children with ADHD contributes to impairments in behavioral inhibition and attentional shifting, leading to difficulties in suppressing responses to irrelevant stimuli and producing abnormally short or prolonged reaction times [ 5 ]. Among executive functions, attentional shifting has emerged as an important cognitive distinction between children with ADHD and typically developing (TD) peers [ 6 – 8 ]. Prior studies have consistently reported that children with ADHD show greater difficulties in inhibition and attentional shifting tasks compared to TD children. Notably, these differences in attentional shifting tend to persist with age [ 6 ]. Traditional assessments of attentional shifting have primarily relied on set-shifting tasks such as the Wisconsin Card Sorting Test and the Trail Making Test-B [ 9 ]. These tasks typically require response changes based on rule shifts or selection among various stimuli, making them more complex than other cognitive tasks, and allowing researchers to estimate attentional shifting ability through measures of accuracy and reaction time (RT). Such complex conditions are necessary because simpler tasks may fail to detect behavioral differences between the ADHD and TD groups [ 10 ]. However, because these tasks engage other high-level cognitive functions, such as working memory, reasoning, and problem-solving, they provide limited insight into attentional shifting at a more fundamental level. Low-level attentional shifting functions as a foundational mechanism for set shifting. Attentional shifting has three sequential stages: orienting to sensory events, detecting signals for focal processing, and maintaining an alert state [ 11 ]. These processes together support the ability to flexibly shift mental sets in response to environmental changes, ultimately involving activation of the prefrontal cortex. However, conventional set-shifting tasks are more suited for assessing integrated executive functioning under cognitively demanding conditions and are therefore less effective for isolating early-stage orienting mechanisms [ 9 , 12 ]. To better understand the behavioral mechanisms underlying complex attentional shifts, it is essential to examine attentional shifting at this more basic level. The gaze-cuing task provides a simplified paradigm for studying low-level attentional shifting by presenting a cue followed by a target. This task has been widely used to assess attentional mechanisms in children with Autism Spectrum Disorder (ASD) [ 13 , 14 ]. This task has also been adapted for ADHD studies, as children with ADHD often demonstrate reduced responsiveness to social cues and frequently exhibit comorbid features with ASD [ 13 , 15 – 18 ]. Most previous studies have focused on behavioral indices, such as key responses, to evaluate whether participants responded to social cues. However, relying solely on key-press responses limits a comprehensive understanding of the attentional shifting process. The introduction of eye-tracking technology into gaze-cueing tasks has enabled more detailed assessments of attentional processes, including eye movements and fixation patterns [ 17 , 18 ]. For instance, several studies have analyzed fixation durations within cue areas to examine gaze responsiveness [ 17 ]. Nevertheless, these studies primarily focus on cue utilization prior to target detection, rather than capturing the attentional shifting that occurs during early-stage orienting behavior in attentional shifts. While some studies have used various tasks, such as memory tasks, to differentiate ADHD, studies specifically targeting attention movement within gaze-cueing tasks remain limited [ 18 ]. Although most previous studies have focused on higher-level attentional shifting within complex tasks [ 6 – 8 ], the present study is the first to reveal subtle delays in low-level attentional shifting in children with ADHD using a simplified gaze-cueing task with minimal complexity. Therefore, this study aims to investigate low-level attentional shifting characteristics in children with ADHD using a simplified gaze-cueing task with reduced distractors. By analyzing both gaze movements and behavioral response data during a straightforward goal-oriented task, we seek to determine whether children with ADHD exhibit slower orienting behavior compared to TD children. The objective of this study is to identify characteristic differences in attentional shifting between children with ADHD and TD children using a low-difficulty gaze-cueing task. Specifically, we aim to determine whether there are group differences in low-level attention shifts under conditions with minimal distractors, assessed through behavioral responses and eye-tracking data. To this end, the following hypotheses were formulated: In a simple gaze-cueing task, eye movement data will more effectively differentiate children with ADHD than key-press-based behavioral data. Under minimal distractor interference, children with ADHD will exhibit less efficient orienting behavior compared to TD children. Children with ADHD will show slower attentional shifting and increased reliance on peripheral vision rather than direct target detection. 2. Materials and methods 2.1. Participants A total of 44 TD children and 28 early elementary school-aged children with ADHD, aged 6–9 years, were initially recruited between August 2021 and February 2022. TD children were recruited through an online community and local flyers, whereas children with ADHD were recruited from Korean medical centers and university hospitals. Eligibility was determined using an online application form that included exclusion criteria to ensure participants met the inclusion requirements. Participants in the ADHD group had recently been diagnosed by a psychiatrist and had been on ADHD medication for less than one year. Those exhibiting overt hyperactivity were excluded, as such behavior could lead to easy identification of ADHD status. Participants with visual acuity below 0.3 were excluded due to potential difficulties in perceiving visual cues during the task. Additional exclusion criteria included diagnoses of other developmental disorders, psychiatric illnesses, congenital genetic conditions, neurological diseases, or a history of acquired brain injury. Further exclusions were made during data processing. Five children diagnosed with both ASD and ADHD were excluded to eliminate potential confounds from comorbid developmental disorders. Additionally, participants with an intelligence quotient below 70, as assessed by the Wechsler Intelligence Scale for Children—Fourth Edition, were excluded to control for borderline intellectual disability. To ensure clearer group distinctions, children in the ADHD group with Korean ADHD Rating Scale (KARS) inattention scores below 6 (n = 1), and children in the TD group with KARS inattention scores above 10 (n = 4), were excluded. Participants were excluded if they failed the eye-tracker calibration (n = 9) or if their average visual deviation exceeded 2° during the validation trial (n = 5), even if initial calibration was successful. Although no sessions were terminated due to participant difficulty, data from one child in each group were lost due to equipment failure and were excluded. Figure 1 presents the overall flow chart. The final sample consisted of 45 children: 19 with ADHD and 26 TD children. This study was approved by the Central Research Facilities Research Ethics Board of the Ulsan National Institute of Science and Technology (UNISTIRB-20-62-A). All research procedures were performed in accordance with the relevant guidelines and regulations, including the Declaration of Helsinki. Written informed consent was obtained from the parents or legal guardians of all participants prior to their participation in the study. 2.2. Measurement Additional exclusion criteria were implemented through self-reported surveys and clinical diagnoses to ensure eligibility in both clinical and nonclinical groups. The KARS was used as a primary screening tool. The scale includes two subscales: inattention and hyperactivity/impulsivity. As the study focused on recruiting children with ADHD who exhibited minimal hyperactivity, inattention scores were used as the primary screening criterion. In a previous study, the average inattention score in the ADHD group was 11.22 (standard deviation [SD] = 5.75). Therefore, ADHD participants with scores 1 SD below the mean (≤ 5) were excluded. TD children were excluded if their inattention score exceeded 10 or their total KARS score surpassed the typical range for their age (≥ 18) [ 19 ]. KARS scores were also used as predictor variables to assess group differences. The Childhood Autism Rating Scale (CARS) was used to screen for autism-related symptoms that might affect performance on the eye-tracking task (Cronbach’s alpha = 0.87). CARS is a widely used screening tool applicable regardless of age or cognitive ability. Participants with a total score of 37 or higher were considered to be at high risk for autism and were excluded from the study [ 20 ]. However, none of the participants exceeded this cutoff. Kovac’s Children’s Depression Inventory (CDI) was used to assess baseline levels of depression [ 21 ], and the State-Trait Anxiety Inventory for Children (STAI-C) was used to measure anxiety [ 22 ]. The CDI, adapted for children aged 8–13, comprises 27 multiple-choice items rated 0 to 2. It comprises 27 multiple-choice items, each rated on a scale from 0 to 2. A total score of 22 or above indicates mild depression, 26 or above indicates moderate depression, and 29 or above indicates severe depression. The Cronbach’s alpha for the CDI in previous studies was 0.898 [ 23 ]. The STAI-C measures anxiety in children using 20 items each for state and trait anxiety. In this study, we used a validated Korean version adapted for elementary school students. Each item is rated on a scale from 1 (not at all) to 3 (often), with higher scores indicating greater anxiety levels (Cronbach’s alpha = 0.892) [ 23 ]. Behavioral data were collected through keyboard inputs, measuring RT for each trial when the participant pressed a key corresponding to the target’s location and accuracy. Eye-tracking data were collected using a Tobii TX 300 eye tracker with a 300 Hz sampling rate. The monitor was 23”, with a screen resolution of 1920 × 1080, and data were collected using the Titta toolbox supported by the Tobii Pro SDK in MATLAB [ 24 ]. Eye movement data were smoothed using the Savitzky–Golay filter (sgolayfilt) with a polynomial order of 9 and a window size of 21. Eye movements with velocities ≥ 30°/s were classified as saccades, and those < 30°/s were classified as fixations. The distance between the participant’s eyes and the monitor was maintained at 65 cm using a chin rest. The cue, target, and distractor were presented at the four corners of the screen, each positioned 0.2 x-units and 0.2-y units from the center, with a diameter of 77 pixels. 2.3. Material-gaze-cueing task To design a task related to joint attention (JA), we adapted Friesen and Kingstone’s gaze-cueing paradigm [ 25 ]. In this task, social stimuli (faces) and nonsocial stimuli (arrows) were used to cue the target either congruently or incongruently (Fig. 2 ). A cue appeared at the center of the screen and then shifted to point to one of the four corners for 500 ms. The target appeared following a randomized stimulus onset asynchrony (SOA) of 250, 500, or 750 ms. In the congruent condition, the cue pointed in the same direction as the target, whereas in the incongruent condition, the cue pointed to a different corner. The target was presented as a red dot, while a distractor (a blue dot) appeared simultaneously at one of the four corners to introduce interference. In the congruent condition, the distractor appeared in a location not aligned with the cue’s direction. In the incongruent condition, the distractor appeared in the direction of the cue to mislead the child’s attention. The target and distractor each subtended 2° of visual angle, with a consistent viewing distance of 65 cm from the screen. To prevent directional bias, the direction of the cue and the locations of the target and distractor were randomized so that each combination appeared with equal probability. The distractor was randomly positioned in one of the three corners of the screen not occupied by the target. The entire task lasted approximately 15 min. There were four blocks, each consisting of 36 trials, and each block took about 9.2 min to complete. Social and nonsocial cues alternated between blocks, and the number of congruent and incongruent trials was evenly distributed within each block. A break of approximately 1 min was provided after the completion of the first two blocks. Before the start of blocks 1 and 3, a 6-point calibration task was performed to set up the eye tracker. Calibration was performed using the Tobii Pro Eye Tracker Manager software ( https://developer.tobiipro.com/eyetrackermanager/etm-installation-information.html ). Following calibration, a validation task was conducted to confirm calibration accuracy. A point was considered valid if its accuracy was within 2° of visual angle. Validation included three points; data were excluded if more than one point failed. Based on this criterion, 10 TD and 2 children with ADHD were excluded. The average validation accuracy among included participants was 1.84° for TD and 1.57° for ADHD children. 2.4. Data analysis 2.4.1. Definition of features This study extracted both behavioral and eye-tracking features. Behavioral features focused on accuracy and RT, which are commonly used metrics in previous studies [ 26 ]. RT was calculated for each stimulus, regardless of response correctness. Accuracy was calculated per condition, and RT metrics were derived using the median RT values from the trials within each condition. Additionally, RT variability, an indicator of sustained attention, was computed as the SD of RT values [ 8 ]. The definitions and descriptions of eye-tracking features are provided in Table 1 in Multimedia Appendix. JA and peripheral vision behaviors were assessed only in blocks involving social stimuli. JA was defined using AOI centered on the cue (that is, the eye stimuli), and the surrounding screen was divided into quadrants, as illustrated in Fig. 3 . A gaze was classified as JA if it shifted from the central cue AOI to the quadrant indicated by the cue during the target detection period (Fig. 2 B-a). If the participant’s gaze remained fixed on the facial stimuli during the target detection period, without active gaze shifts toward the target location, this behavior was interpreted as target detection via “peripheral vision” (Fig. 2 B-b). To reduce the impact of outliers, continuous variables (except those capturing behavior occurrence) were computed using median values rather than mean values. For instance, the RT metric for social cues in the congruent condition at an SOA of 0.5 s was calculated as the median RT value for that specific condition. For binary variables (0 or 1), e.g., JA presence or peripheral vision usage, mean values represented the proportion of trials exhibiting each behavior. The gaze-cueing effect (GCE) was computed by subtracting the values for congruent stimuli from those for incongruent stimuli under each SOA condition during the target detection phase, in trials with social stimuli [ 26 ]. For example, GCE in RT at an SOA of 0.5 s was calculated by subtracting the RT under the congruent condition from that under the incongruent condition, both for social cues at that specific SOA. 2.4.2. Statistical analysis To assess the appropriateness of the participant pool and identify group-level differences, sociodemographic and clinical characteristic data were analyzed using the Mann–Whitney U rank test, a nonparametric method used to compare the distributions of variables between ADHD and TD groups [ 27 ]. This test was employed to identify features with statistically significant differences between the groups. To correct for multiple comparisons, the Benjamini—Hochberg procedure was applied using the multipletest function from the Statsmodels library in Python [ 28 ]. Only features that remained significant after corrections were retained for further analysis. Next, stepwise logistic regression was conducted for feature selection using the StepReg library in R Studio [ 29 ]. This method employed bidirectional elimination, where variables were included based on their statistical significance. The entry and exit criteria for variable selection were based on Rao’s score test, with the significance threshold at 0.035. Both behavioral and eye-tracking features were subjected to this selection process, and the features identified as significant were used as inputs to train a combined predictive model. Logistic regression modeling was then performed using selected features. To mitigate the risk of overfitting due to the small sample size, fivefold cross-validation was applied. Modeling was conducted in R Studio using the mlr3 package, and cross-validation was repeated 50 times to improve generalizability and reduce bias. During each iteration, label distribution was preserved across folds to maintain the balance between the ADHD and TD groups. For features found to be insignificant during modeling, feature importance was evaluated using the varImp function from the caret library in R Studio. Features with low importance scores were iteratively removed to determine whether this improved the performance of the model, particularly its area under the curve (AUC). This pruning process continued until no further AUC improvement was observed, resulting in the final feature set and model. Model performance was evaluated based on the average accuracy, AUC, and F1 score (which incorporates both precision and recall) across the five validation folds. Consistency of the five cross-validation iterations was assessed using Cohen’s kappa coefficient, which measures the level of agreement between the actual labels and the predicted labels. Cohen’s kappa values were interpreted as follows: 0: Poor agreement 0–0.2: Slight agreement 0.21–0.40: Fair agreement 0.41–0.60: Moderate agreement 0.61–0.80: Substantial agreement 0.81–1.00: Perfect agreement [ 30 ]. 3. Results 3.1. Sociodemographic and clinical characteristics of participants The ADHD and TD groups were comparable in age, as both consisted of early elementary school–aged children. Although there was a difference in sex distribution, this disparity was expected due to the higher prevalence of ADHD diagnoses in males. The KARS, which measures ADHD symptom severity, revealed significant group differences in both the inattention and hyperactivity/impulsivity subscales. Additionally, scores on the CARS differed significantly, supporting prior research indicating that children with ADHD tend to exhibit more autistic traits than their TD peers [ 31 ]. No significant differences were found between the groups in state or trait anxiety. However, depression scores were significantly higher in the ADHD group (p = 0.003). Detailed sociodemographic information is summarized in Table 2 . Table 1 Sociodemographic and clinical characteristics of each group. Measures (mean \(\:\pm\:\) SD ) ADHD (n = 19) TD (n = 27) P value Age 7.7 \(\:\pm\:\) 1.0 7.9 \(\:\pm\:\) 0.9 0.709 Sex (male n, %) 16 (84.2) 12 (44.4) - KARS total score 22.4 \(\:\pm\:\) 11.2 \(\:5.6\pm\:\) 4.4 < 0.001 Inattention score 13.7 \(\:\pm\:\) 6.0 3.5 \(\:\pm\:\) 2.8 < 0.001 Hyperactivity/impulsivity score 8.7 \(\:\pm\:\) 6.0 2.1 \(\:\pm\:\) 2.1 < 0.001 CARS 18.7 \(\:\pm\:\) 5.3 15.1 \(\:\pm\:\) 0.4 < 0.001 CDI 13.0 \(\:\pm\:\) 6.5 7.7 \(\:\pm\:\) 3.7 0.003 STAI-C Trait anxiety 30.3 \(\:\pm\:\) 5.8 28.3 \(\:\pm\:\) 5.2 0.259 State anxiety 33.3 \(\:\pm\:\) 6.9 30.6 \(\:\pm\:4.9\) 0.176 3.2. Hypothesis 1: During a simple gaze-cueing task, eye movement data can distinguish children with ADHD more effectively than behavioral data The Mann–Whitney U test was used to compare group differences across measured variables. As summarized in Table 2 , trials using social cues showed significant group differences in the number of saccades, saccade velocity, and the SD of fixation locations. Approximately half of the significantly different features were related to JA. The frequency of JA was higher in the TD group compared to the ADHD group. Conversely, the difference in the number of saccades between incongruent and congruent conditions was larger in the ADHD group, with this contrast reaching high significance (p < .001). In trials utilizing nonsocial cues, significant indicators included the number of saccades, saccade velocity, SD of fixation locations, saccade length, and the ratio of null data (Table 2 ). Table 2 Distribution of eye-tracking features between ADHD and TD groups, selected using the Mann–Whitney U test. Stimuli Data sort Feature (SOA/congruency) ADHD TD Cohen’s D p-value Social Eye movement Number of saccades (0.25 s/cong) 6.89 ± 6.03 10.3 ± 5.5 -0.594 0.043 Number of saccades (0.5 s/cong) 5.89 ± 4.33 10.46 ± 6.09 -0.84 0.037 Number of saccades (0.75 s/cong) 6.66 ± 5.57 9.69 ± 5.49 -0.548 0.043 Saccade velocity (0.5 s/incong) 89.56 ± 22.18 100.09 ± 24.45 -0.447 0.043 Saccade velocity (0.75 s/incong) 89.78 ± 21.21 100.08 ± 25.84 -0.428 0.045 SD of fixation locations (0.25 s/cong) 4.91 ± 3.16 6.29 ± 3.1 -0.443 0.045 Rate of JA (0.25 s/cong) 0.24 ± 0.14 0.34 ± 0.1 -0.804 0.043 Rate of JA (0.25 s/incong) 0.2 ± 0.13 0.3 ± 0.11 -0.834 0.037 GCE in the saccade number (0.5 s/incong–cong) 3.13 ± 2.99 -0.57 ± 3.02 1.233 0.009 Rate of peripheral vision (0.25 s/incong) 0.64 ± 0.16 0.56 ± 0.11 0.627 0.046 Behavior GCE in RT (0.5 s/incong–cong) 0.12 ± 0.15 0.03 ± 0.07 0.855 0.045 Non- social Eye movement Saccade length (0.75 s/cong) 11.3 ± 6.02 16.59 ± 8.32 -0.708 0.041 Saccade length (0.75 s/incong) 20.12 ± 13.15 27.02 ± 10.67 -0.588 0.043 Number of saccades (0.25 s/cong) 5.71 ± 4.53 9.22 ± 5.5 -0.685 0.037 Number of saccades (0.25 s/incong) 6.61 ± 4.72 10.81 ± 6.28 -0.739 0.037 Saccade velocity (0.5 s/incong) 96.97 ± 23.09 105.08 ± 21.53 -0.366 0.043 Saccade velocity (0.75 s/cong) 83.94 ± 18.39 94.76 ± 19.11 -0.575 0.043 Saccade velocity (0.75 s/incong) 93.67 ± 20.15 104.71 ± 23.69 -0.495 0.037 SD of fixation locations (0.25 s/incong) 7.13 ± 3.46 9.08 ± 3.11 -0.596 0.043 SD of fixation locations (0.5 s/incong) 7.56 ± 3.43 9.43 ± 3.3 -0.557 0.043 SD of fixation locations (0.75 s/incong) 6.39 ± 3.76 9.3 ± 3.51 -0.805 0.037 Rate of null data (0.5 s/cong) 0.07 ± 0.13 0.0 ± 0.02 0.841 0.041 Rate of null data (0.75 s/cong) 0.08 ± 0.1 0.01 ± 0.06 0.808 0.043 Rate of null data (0.75 s/incong) 0.05 ± 0.09 0.0 ± 0.0 0.801 0.043 Behavior Accuracy (0.25 s/incong) 0.76 ± 0.35 0.98 ± 0.03 -0.981 0.037 Total Eye movement Rate of null data 0.11 ± 0.1 0.05 ± 0.04 0.957 0.041 SD of null rate 0.15 ± 0.09 0.1 ± 0.06 0.727 0.043 Behavior SD of RT 0.55 ± 0.52 0.26 ± 0.17 0.818 0.043 Overall, the TD group exhibited more frequent and faster eye movements during the target detection period, along with greater variability in gaze behavior. Three behavioral features were significantly different between the groups. Accuracy under the nonsocial cue condition at an SOA of 0.25 s during incongruent trials differed significantly (p = 0.008), whereas accuracy under other conditions did not. The SD of RT differed significantly across all conditions (p = 0.033), and the GCE in RT at an SOA of 0.5 s also showed a significant group difference (p = 0.044). Stepwise regression analysis identified accuracy and GCE in RT as significant predictors in the behavioral model. For the eye movement model, key features included the number of saccades, GCE in number of saccades, JA rate, and the proportion of null data. Notably, approximately half of the selected features in both models were related to JA. Irrespective of the data source, the model included the SD of the RT, number of saccades, GCE in the number of saccades, and SD of fixation locations when considering the entire dataset. Model performance metrics derived from logistic regression using these features are presented in Table 3 . Table 3 Features used for logistic regression and model performance by each dataset. Dataset Features used for the model Kappa Acc F1 score (Precision/Recall) AUC Behavior - Accuracy (0.25s/incong) * - GCE in RT (0.5) * .435 .741 .788 (.746/.871) .828 Eye movement - Number of saccade (0.25s/incong) * - GCE in number of saccade (0.5s) * - Rate of JA (0.25s/cong) - Rate of null data * .644 .837 .853 (.863/.872) .929 Total - SD of RT * - Number of saccade (0.25s/cong) * - GCE in number of saccade (0.5s) * - SD of fixation locations (0.25s/incong) · .716 .870 .882 (.883/.901) .911 * indicates significant variables in the logistic regression (p < 0.05) The eye movement model outperformed the behavioral model in both accuracy (0.837 vs. 0.741) and AUC (0.929 vs. 0.828). While the accuracy and AUC of the eye movement model were comparable to those of the combined model, the kappa statistic was higher for the integrated model that incorporated both behavioral and eye-tracking data (kappa = 0.716). The behavior-only model demonstrated a moderate level of agreement (k = 0.435), whereas both the eye movement-only and combined models achieved significant agreement according to established kappa interpretation guidelines [ 30 ]. 3.3. Hypothesis 2 : Under minimal distractor interference, children with ADHD show less efficient goal-oriented attentional shifting compared to TD children Although RT did not differ significantly between the groups (p = 0.707), a trend emerged suggesting that the TD group produced a higher number of saccades (p = 0.057, d= -0.516). This trend prompted an examination of fixation duration per saccade, calculated by dividing total fixation duration by the number of saccades during target detection. Results showed that children with ADHD had significantly longer fixation duration per saccade compared to TD children (p = 0.017, d = 0.799), indicating less efficient gaze shifting. However, accuracy was lower in the ADHD group compared to the TD group (p = 0.086). Figure 3 illustrates the distributions of these features. Following the inclusion of eye movement metrics in the subset selection process, model performance remained consistent regardless of whether features were selected from eye-tracking data alone or in combination with behavioral data. The key features identified included: (1) GCE in number of saccades at SOA = 0.5 s, (2) rate of null data under nonsocial cues (SOA = 0.5 s, incongruent), and (3) fixation duration per saccade in social cue trials. Notably, the “fixation duration per saccade” was a novel metric identified specifically under social cue conditions during the target detection period. Incorporating this metric improved the performance of the logistic regression model. The updated model achieved an AUC of 0.940 and an accuracy of 0.860, demonstrating a significant improvement over the previous version (Table 4 ). Table 4 Comparison of the model with and without duration in each fixation. Duration in each fixation - Duration in each fixation + \(\:\varvec{\beta\:}\) Z (p-value) \(\:\varvec{\beta\:}\) Z (p-value) Features SD of RT 14.771 2.467 (.014) Number of saccade (0.25 s/cong) − .842 -2.348 (.019) SD of fixation locations (0.25 s/incong) − .666 -1.676 (.094) GCE in number of saccade (0.5 s) 1.901 2.177 (.030) 1.306 2.485 (.013) null_rate in nonsocial (0.5 s/cong) 114.266 2.449 (.014) fixation duration per saccade 43.157 2.646 (.008) Kappa .716 0.696 Acc .870 .860 F1 score (precision/recall) .882 (.883/.901) 0.876 (0.879/0.891) AUC .911 0.940 The correlations between selected model features and the ADHD symptom severity, as measured by the KARS, were evaluated (Table 2 in Multimedia Appendix). Fixation duration per saccade exhibited the strongest correlation with inattention (r = 0.46, p < .01) and hyperactivity scores (r = 0.36, p < .01). Features such as GCE in number of saccades at an SOA of 0.5 s and the rate of null data in nonsocial cues at an SOA of 0.5 s under incongruent conditions showed a significant correlation with the inattention KARS scores (p < .01). Conversely, among other features, only the SD of RT showed a significant correlation with the hyperactivity KARS scores (r = 0.32, p < .01), excluding the duration at each fixation point. Other features included in the model did not show statistically significant correlations with KARS scores (Table 2 in Multimedia Appendix). 3.4. Hypothesis 3: Children with ADHD exhibit slower attentional shifting and rely more on peripheral vision rather than actively detecting the target. In trials involving social cues, the rate of JA was significantly higher in the TD group than in the ADHD group (p = 0.042, d = -0.827), regardless of SOA or cue congruency. Furthermore, despite comparable RTs, the ADHD group demonstrated significantly longer fixation durations per saccade than the TD children (p = 0.017, d = 0.799). Figure 4 shows representative eye movement patterns for the two groups. These findings suggest that children with ADHD tend to restrict their gaze to the AOI surrounding the central cue, rather than executing goal-directed gaze shifts. Additionally, the ADHD group exhibited a significantly higher frequency of peripheral vision use during the target detection period compared to the TD group, as illustrated in Fig. 3 (p = 0.046, d = 0.481). 4. Discussion This study utilized a gaze-cuing task to collect both behavioral and eye movement data from children, with the aim of identifying indicators of ADHD and developing an early-stage classification model. Our findings supported all three hypotheses and provided new insights into low level of attentional and executive functioning deficits in children with ADHD, even in low-demand tasks. First, consistent with our initial hypothesis, this study highlights the diagnostic utility of eye movement data in detecting ADHD, particularly in low-demand tasks where behavioral differences are subtle. Unlike previous studies that reported significant differences in accuracy among children with ADHD [ 32 ], we observed significant behavioral differences in accuracy only under the social cue condition with a SOA of 0.25 s and incongruent cues. Conversely, the model developed in this study achieved relatively high performance using only eye movement indicators (AUC = 0.929, accuracy = 0.837). These results suggest that the gaze-cueing task employed in this study may not have been sufficiently demanding to differentiate the groups based on behavioral indicators alone [ 31 ]. Second, our findings support the hypothesis that children with ADHD exhibit less efficient goal-oriented attentional shifting, even in environments with minimal distractor interference. This was demonstrated by prolonged fixation durations and reduced saccadic frequencies during the target detection behavior. These eye movement patterns suggest delayed attentional disengagement from irrelevant areas and reduced engagement in active visual search—both of which are essential for effective attentional shifting. These findings align with prior research linking reduced saccadic frequency to impaired attentional control and suggesting that limited exploratory eye movements reflect executive dysfunction in ADHD [ 33 ]. Moreover, fixation duration per saccade during target detection emerged as the most predictive feature in our classification model, showing strong correlations with KARS inattention (r = 0.46, p < 0.01) and hyperactivity scores (r = 0.36, p < 0.01). Prior meta-analyses have indicated that small to moderate deficits in attentional shifting are associated with inattention and hyperactivity [ 34 ]. Our findings further suggest that prolonged fixations reflect more than motor slowing—they may indicate a reduced capacity for flexible attentional engagement. While previous work has shown that ADHD may not impair set-shifting speed when high-level executive functions remain intact [ 35 ], our findings suggest deficits at more automatic, early stages of orienting. This supports a hierarchical model of attentional control, where ADHD symptoms manifest at multiple cognitive levels. The subtle nature of these low-level impairments highlights the importance of examining nontraditional indicators, such as prolonged fixations, to capture executive dysfunction beyond what standard neuropsychological tasks can detect. Third, our data support the hypothesis that children with ADHD rely more on peripheral vision than on active, intentional target detection. Compared with TD children, the ADHD group exhibited both a lower rate of JA and a higher rate of peripheral vision use. JA serves as a regulatory mechanism that facilitates shared attention and social coordination [ 17 , 36 ]. A reduced rate of JA in the ADHD group suggests impaired use of external cues for attentional guidance. This deficit may indicate overlapping impairments in social cue responsiveness, consistent with the high comorbidity between ADHD and ASD [ 37 ]. Given that reduced JA is a prominent early marker for ASD, our finding of diminished JA in children with ADHD supports the notion of shared difficulties in processing socially relevant cues [ 38 ]. Although a previous study reported increased face fixation time in children with ADHD across the task duration [ 17 ], this study adopts a more focused approach by examining eye movements immediately following gaze cue onset. This allowed us to assess the visual strategies employed specifically during goal-directed target detection. Our findings reveal that children with ADHD are more likely to rely on peripheral visual input and to engage in less structured visual scanning. Notably, the ADHD group demonstrated a greater difference in saccade frequency between congruent and incongruent trials than the TD group. This pattern may reflect heightened sensitivity to conflicting cues, indicating reduced cognitive control in situations that demand attentional flexibility [ 35 ] Across all three hypotheses, eye movement patterns in the ADHD group consistently indicated delayed and inefficient attentional shifting, as well as difficulty using cues to guide attention. These impairments—evident even in a distraction-free, low-demand task—highlight fundamental executive dysfunctions in ADHD patients and suggest that deficits in attentional shifting impairments may serve as robust behavioral markers for early identification. This study presents several notable strengths. First, it employed a simple gaze-cueing task devoid of complex distractors to investigate low-level attentional shifting in children with ADHD. Although such paradigms are commonly used in developmental research, they have rarely been applied to detect subtle attention deficits in ADHD populations. Despite minimal task demands, significant group differences emerged in eye-tracking measures. Second, the study identified prolonged fixation duration during target detection as a novel and highly predictive marker of ADHD. This feature was linked to reduced attentional shifting efficiency and showed strong correlations with clinical indices of inattention and hyperactivity. Its inclusion in the classification model significantly improved predictive accuracy, underscoring its potential as a core feature for early ADHD detection. Furthermore, by analyzing eye movement patterns, this study provided insight into the behavioral mechanisms underlying attentional shifts in ADHD. These findings revealed physiological indicators that may facilitate early identification, and could be integrated with other bio-signals—such as electroencephalography—to enhance diagnostic precision in future multimodal assessments. Lastly, despite the limited sample size, the study achieved high classification accuracy and AUC values using a minimal set of eye movement features. The use of cross-validation techniques helped reduce overfitting, suggesting these indicators may scale well in larger diagnostic frameworks. Nevertheless, some limitations should be acknowledged. The relatively low task complexity may have reduced behavioral differentiation between groups, although eye movement data effectively compensated for this limitation. Additionally, the small sample size may restrict the generalizability of the findings. Future studies with larger and more diverse samples are needed to validate and refine these findings. This study demonstrates that even a simple gaze-cueing task, analyzed using a logistic regression model based on behavioral and eye movement data, can objectively detect atypical attentional shifting in children with ADHD. Notably, eye-tracking indicators, particularly prolonged fixation durations per saccade and reduced saccade frequency, proved more sensitive than traditional behavioral responses in distinguishing ADHD, emerging as strong markers linked to core symptoms. In particular, fixation duration during target detection showed significant correlations with inattention and hyperactivity. These findings underscore the potential of combining simplified gaze-cueing tasks with eye-tracking technology to reveal subtle executive dysfunctions and enable early, efficient ADHD screening. The strong classification performance further supports the utility of these markers. Future research should expand sample sizes and explore these markers across diverse populations and settings to enhance generalizability and refine diagnostic tools. Declarations Author Contributions: L.S.M. conceived the study, collected data, performed preprocessing and statistical analyses, and wrote the original draft. L.S.I. contributed to the conceptualization, experimental design, discussion, and revision of the manuscript. J.I.J. collected data and contributed to the original draft. J.J.H. collected data and contributed to data preprocessing and statistical analyses. P.H.J. contributed to the conceptualization. K.M.K. contributed to the conceptualization, provided advice on data analysis, revised the manuscript, and acquired funding. Z.T. contributed to the discussion and revision of the manuscript. S.S. contributed to the conceptualization and funding acquisition. J.D.Y. contributed to the conceptualization and methodology, revised the manuscript, supervised the project, and acquired funding. All authors reviewed and approved the final manuscript. Institutional Review Board Statement: This study was approved by the Central Research Facilities Research Ethics Board of the Ulsan National Institute of Science and Technology (UNISTIRB-20-62-A). Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: The datasets generated and/or analyzed during this study are not publicly available due to participant confidentiality agreements, but are available from the corresponding author on reasonable request. Funding: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2020M3E5D9080787). This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (Grant Number: HI22C0646). This study was supported by the U-K (UNIST-Korea) research brand program (1.230016.01) funded by UNIST (Ulsan National Institute of Science & Technology). Conflicts of Interest: The authors have no conflict of interest to declare. References Biederman, J. Attention-deficit/hyperactivity disorder: a selective overview. Biol. Psychiatry . 57 , 1215–1220 (2005). Faraone, S. V., Biederman, J. & Mick, E. The age-dependent decline of attention deficit hyperactivity disorder: a meta-analysis of follow-up studies. Psychol. Med. 36 , 159–165 (2006). Pennington, B. F. & Ozonoff, S. Executive functions and developmental psychopathology. J. Child. Psychol. Psychiatry . 37 , 51–87 (1996). Brown, S. W. Timing and executive function: Bidirectional interference between concurrent temporal production and randomization tasks. Mem. Cognit . 34 , 1464–1471 (2006). Barkley, R. A. Behavioral inhibition, sustained attention, and executive functions: constructing a unifying theory of ADHD. Psychol. Bull. 121 , 65 (1997). Qian, Y. et al. The developmental trajectories of executive function of children and adolescents with attention deficit hyperactivity disorder. Res. Dev. Disabil. 34 , 1434–1445 (2013). Çak, H. T. et al. The behavior rating inventory of executive function and continuous performance test in preschoolers with attention deficit hyperactivity disorder. Psychiatry Investig . 14 , 260–265 (2017). Shalev, L. et al. Conjunctive Continuous Performance Task (CCPT)—A pure measure of sustained attention. Neuropsychologia 49 , 2584–2591 (2011). Boshomane, T. T., Pillay, B. J. & Meyer, A. Mental flexibility (set-shifting) deficits in children with ADHD: a replication and extension study. J. Psychol. Afr. 31 , 344–349 (2021). Huizenga, H. M. et al. Task complexity enhances response inhibition deficits in childhood and adolescent attention-deficit/hyperactivity disorder: a meta-regression analysis. Biol. Psychiatry . 65 , 39–45 (2009). Posner, M. I. & Petersen, S. E. The attention system of the human brain. Annu. Rev. Neurosci. 13 , 25–42 (1990). Sherigar, S. S., Gamsa, A. H. & Srinivasan, K. Oculomotor deficits in attention deficit hyperactivity disorder: a systematic review and meta-analysis. Eye 37 , 1975–1981 (2023). Seernani, D. et al. Social and non-social gaze cueing in autism spectrum disorder, attention-deficit/hyperactivity disorder and a comorbid group. Biol. Psychol. 162 , 108096 (2021). Stallworthy, I. C. et al. Variability in responding to joint attention cues in the first year is associated with autism outcome. J. Am. Acad. Child. Adolesc. Psychiatry . 61 , 413–422 (2022). Temeltürk, R. D. et al. Dynamic eye-tracking evaluation of responding joint attention abilities and face scanning patterns in children with attention deficit hyperactivity disorder. Dev. Psychopathol. 36 , 1190–1201 (2024). Marotta, A. et al. Impaired reflexive orienting to social cues in attention deficit hyperactivity disorder. Eur. Child. Adolesc. Psychiatry . 23 , 649–657 (2014). Temeltürk, R. D. et al. Dynamic eye-tracking evaluation of responding joint attention abilities and face scanning patterns in children with attention deficit hyperactivity disorder. Dev. Psychopathol. 36 , 1–12 (2023). Yoo, J. H. et al. Development of an innovative approach using portable eye tracking to assist ADHD screening: a machine learning study. Front. Psychiatry . 15 , 1337595 (2024). Jang, S. J., Suh, D. S. & Byun, H. J. Normative study of the K-ARS (Korean ADHD Rating Scale) for parents. J. Korean Acad. Child. Adolesc. Psychiatry . 18 , 38–48 (2007). Lee, N., Hong, Y. & Kim, Y. Screening instruments for autism spectrum disorder: mini review. J. Korean Neuropsychiatr Assoc. 58 , 192–201 (2019). Cho, S. & Lee, Y. Development of the Korean form of the Kovacs’ Children’s Depression Inventory. J. Korean Neuropsychiatr Assoc. 29 , 943–956 (1990). Choi, J. & Cho, S. Assessment of anxiety in children: Reliability and validity of revised children’s manifest anxiety scale. J. Korean Neuropsychiatr Assoc. 29 , 691–702 (1990). Lee, W. J. et al. The propensity to depression and anxiety in children. J. Korean Acad. Fam Med. 19 , 828–837 (1998). Niehorster, D. C., Andersson, R. & Nyström, M. Titta: a toolbox for creating PsychToolbox and Psychopy experiments with Tobii eye trackers. Behav. Res. Methods . 52 , 1970–1979 (2020). Friesen, C. K. & Kingstone, A. The eyes have it! Reflexive orienting is triggered by nonpredictive gaze. Psychon Bull. Rev. 5 , 490–495 (1998). Shin, W. G. et al. Individual differences in gaze-cuing effect are associated with facial emotion recognition and social conformity. Front. Psychol. 14 , 1219488 (2023). MacFarland, T. W. & Yates, J. M. Introduction to Nonparametric Statistics for the Biological Sciences Using R (Springer, 2016). Benjamini, Y. & Hochberg, Y. On the adaptive control of the false discovery rate in multiple testing with independent statistics. J. Educ. Behav. Stat. 25 , 60–83 (2000). Li, J. et al. Package ‘StepReg’ . (2020). Wardhani, N. W. S. et al. IEEE,. Cross-validation metrics for evaluating classification performance on imbalanced data. In Proc. 2019 Int. Conf. Comput. Control Inform. Appl. (IC3INA) 45–50 (2019). Hattori, J. et al. Are pervasive developmental disorders and attention-deficit/hyperactivity disorder distinct disorders? Brain Dev. 28 , 371–374 (2006). Jones, D. R. et al. Prevalence, severity, and co-occurrence of chronic physical health problems of persons with serious mental illness. Psychiatr Serv. 55 , 1250–1257 (2004). Lambek, R. et al. Executive dysfunction in school-age children with ADHD. J. Atten. Disord . 15 , 646–655 (2011). Willcutt, E. G. et al. Neuropsychological analyses of comorbidity between reading disability and attention deficit hyperactivity disorder: In search of the common deficit. Dev. Neuropsychol. 27 , 35–78 (2005). Irwin, L. N. et al. Do children with attention-deficit/hyperactivity disorder (ADHD) have set shifting deficits? Neuropsychology 33 , 470–481 (2019). Drigas, A. & Karyotaki, M. Attentional control and other executive functions. Int. J. Emerg. Technol. Learn. 12 , 219–230 (2017). Mundy, P. et al. Brief report: joint attention and information processing in children with higher functioning autism spectrum disorders. J. Autism Dev. Disord . 46 , 2555–2560 (2016). Hours, C., Recasens, C. & Baleyte, J. M. ASD and ADHD comorbidity: What are we talking about? Front. Psychiatry . 13 , 837424 (2022). Additional Declarations No competing interests reported. Supplementary Files Appendix1.pdf Cite Share Download PDF Status: Published Journal Publication published 26 Feb, 2026 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 12 Nov, 2025 Reviews received at journal 08 Oct, 2025 Reviews received at journal 19 Sep, 2025 Reviewers agreed at journal 17 Sep, 2025 Reviewers agreed at journal 12 Sep, 2025 Reviews received at journal 11 Sep, 2025 Reviewers agreed at journal 01 Sep, 2025 Reviewers invited by journal 20 Aug, 2025 Editor invited by journal 29 Jul, 2025 Editor assigned by journal 28 Jul, 2025 Submission checks completed at journal 26 Jul, 2025 First submitted to journal 26 Jul, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7191610","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":504692447,"identity":"08fba913-1b4c-41fc-a480-4e3992aab269","order_by":0,"name":"Seonmi Lee","email":"","orcid":"","institution":"Ulsan National Institute of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Seonmi","middleName":"","lastName":"Lee","suffix":""},{"id":504692448,"identity":"0ede8303-8e0a-4f1b-9c80-b8bc8aeca1c3","order_by":1,"name":"Sangil Lee","email":"","orcid":"","institution":"Ulsan National Institute of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Sangil","middleName":"","lastName":"Lee","suffix":""},{"id":504692449,"identity":"90bd7893-b804-47c0-a93f-bcf1c14fcf1d","order_by":2,"name":"Inji Jeong","email":"","orcid":"","institution":"Ulsan National Institute of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Inji","middleName":"","lastName":"Jeong","suffix":""},{"id":504692450,"identity":"1a8ebe9d-7dfb-4619-86fb-540db8a9b1c0","order_by":3,"name":"Jaehyun Jeong","email":"","orcid":"","institution":"LVIS Korea","correspondingAuthor":false,"prefix":"","firstName":"Jaehyun","middleName":"","lastName":"Jeong","suffix":""},{"id":504692451,"identity":"1b9ee6eb-acc5-4fa4-b99a-bdb258072dd1","order_by":4,"name":"Hyoju Park","email":"","orcid":"","institution":"Pusan National University","correspondingAuthor":false,"prefix":"","firstName":"Hyoju","middleName":"","lastName":"Park","suffix":""},{"id":504692452,"identity":"6caf4425-5e15-46b7-ba57-9d845e8f1f20","order_by":5,"name":"Mee-Kyoung Kwon","email":"","orcid":"","institution":"Seoul Women’s University","correspondingAuthor":false,"prefix":"","firstName":"Mee-Kyoung","middleName":"","lastName":"Kwon","suffix":""},{"id":504692453,"identity":"d20c5634-5b9a-4f02-9db3-6be4a25692d4","order_by":6,"name":"Theodore Zanto","email":"","orcid":"","institution":"University of California","correspondingAuthor":false,"prefix":"","firstName":"Theodore","middleName":"","lastName":"Zanto","suffix":""},{"id":504692454,"identity":"dcc34d7d-0a33-4bca-92d8-dd71d53db184","order_by":7,"name":"Sunhae Sul","email":"","orcid":"","institution":"Pusan National University","correspondingAuthor":false,"prefix":"","firstName":"Sunhae","middleName":"","lastName":"Sul","suffix":""},{"id":504692455,"identity":"785b8aae-18e4-4782-8064-de277c6b570f","order_by":8,"name":"Dooyoung Jung","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA30lEQVRIiWNgGAWjYDCCA1CaH8FmI1KLZAOInUCKFgMwgxgtfMebj0l8bLtjt/lG7sGDP38wyPM3sKV9wKdF8syxNMmZbc+St93ISzjMk8BgOOMA2+EZ+LQY3Mgxk+ZtO5xsdiPH4DDQYYwbGNib8ToMrsV4Ro7BwR8JDPZEa7EzkMgxOAB0WOIGBrbDeLUA/ZJsOePc4QSJM28MDvOkSSTPOMyWjFcLMMQO3vhQdtievz3H+OMPGxvb/vY2Y7xagIBFAkgkNkA4QDYzIQ1AJaBYsCesbhSMglEwCkYsAABd6Ew3hdSEdwAAAABJRU5ErkJggg==","orcid":"","institution":"Ulsan National Institute of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Dooyoung","middleName":"","lastName":"Jung","suffix":""}],"badges":[],"createdAt":"2025-07-23 03:23:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7191610/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7191610/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-026-41419-0","type":"published","date":"2026-02-26T15:58:27+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":90306168,"identity":"4d74db08-fa9f-4b23-a5d8-916fc33c9981","added_by":"auto","created_at":"2025-09-01 09:26:30","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":528038,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart of study participants and eligibility.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-7191610/v1/c51b09e3da8f0030e3ea98cb.png"},{"id":90307769,"identity":"12d00a44-aa8e-4172-9278-e501853a19a0","added_by":"auto","created_at":"2025-09-01 09:34:30","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":439554,"visible":true,"origin":"","legend":"\u003cp\u003eOverview of Experimental Task Involving Joint Attention and Peripheral Vision. A Sequence of gaze-cueing task. B-(a) AOI to determine JA B-(b) peripheral vision occurrence.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-7191610/v1/cd91e093d475e5951e8efdd9.png"},{"id":90306174,"identity":"da5c17b8-25a9-4aaf-9a14-9055ba53c2ef","added_by":"auto","created_at":"2025-09-01 09:26:30","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":77592,"visible":true,"origin":"","legend":"\u003cp\u003eFeature exploration during target detection with social cues.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-7191610/v1/d180fcf45ce375f06307c068.png"},{"id":90310207,"identity":"de3a499d-f10c-4979-a004-84347a1c3455","added_by":"auto","created_at":"2025-09-01 09:42:31","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":405954,"visible":true,"origin":"","legend":"\u003cp\u003eEye movement pathways from cue to target detection across 10 trials. (a) Eye movement trajectory of one participant from the TD group during 10 trials of the gaze-cuing task; (b) Eye movement trajectory of one participant from the ADHD group during the same task.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-7191610/v1/8d144899662940d3d1d30d09.png"},{"id":103765571,"identity":"435091c3-163a-4eb3-a756-834c5ea94498","added_by":"auto","created_at":"2026-03-02 16:04:44","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2621641,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7191610/v1/2f160a68-e304-4715-860c-bbb29295c99b.pdf"},{"id":90306169,"identity":"0d3299da-52a0-4871-80bc-2488f24f7c0b","added_by":"auto","created_at":"2025-09-01 09:26:30","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":179758,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7191610/v1/beb3edb1d3a0c8f97135c35a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Exploring early-stage orienting behavior using an eye tracker for attention deficit hyperactivity disorder classification","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eAttention deficit hyperactivity disorder (ADHD) is a prototypical neurodevelopmental disorder characterized by persistent symptoms of inattention, hyperactivity, and impulsivity. Although it typically manifests during childhood, these symptoms often persist into adolescence and adulthood [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. ADHD significantly impacts academic performance and daily functioning, primarily due to difficulties in sustaining attention, and is closely associated with various cognitive deficits.\u003c/p\u003e\u003cp\u003eTheoretically, deficits in attention and impulsivity associated with ADHD have been attributed to impairments in executive functions [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Executive function refers to higher-order cognitive processes that regulate goal-directed behavior by integrating stimuli and controlling responses. These processes include stimulus prioritization, response inhibition, and cognitive shifting to achieve specific goals [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Barkley (1997) notably suggested that reduced prefrontal cortex activity in children with ADHD contributes to impairments in behavioral inhibition and attentional shifting, leading to difficulties in suppressing responses to irrelevant stimuli and producing abnormally short or prolonged reaction times [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAmong executive functions, attentional shifting has emerged as an important cognitive distinction between children with ADHD and typically developing (TD) peers [\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Prior studies have consistently reported that children with ADHD show greater difficulties in inhibition and attentional shifting tasks compared to TD children. Notably, these differences in attentional shifting tend to persist with age [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTraditional assessments of attentional shifting have primarily relied on set-shifting tasks such as the Wisconsin Card Sorting Test and the Trail Making Test-B [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. These tasks typically require response changes based on rule shifts or selection among various stimuli, making them more complex than other cognitive tasks, and allowing researchers to estimate attentional shifting ability through measures of accuracy and reaction time (RT). Such complex conditions are necessary because simpler tasks may fail to detect behavioral differences between the ADHD and TD groups [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. However, because these tasks engage other high-level cognitive functions, such as working memory, reasoning, and problem-solving, they provide limited insight into attentional shifting at a more fundamental level.\u003c/p\u003e\u003cp\u003eLow-level attentional shifting functions as a foundational mechanism for set shifting. Attentional shifting has three sequential stages: orienting to sensory events, detecting signals for focal processing, and maintaining an alert state [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. These processes together support the ability to flexibly shift mental sets in response to environmental changes, ultimately involving activation of the prefrontal cortex. However, conventional set-shifting tasks are more suited for assessing integrated executive functioning under cognitively demanding conditions and are therefore less effective for isolating early-stage orienting mechanisms [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. To better understand the behavioral mechanisms underlying complex attentional shifts, it is essential to examine attentional shifting at this more basic level.\u003c/p\u003e\u003cp\u003e The gaze-cuing task provides a simplified paradigm for studying low-level attentional shifting by presenting a cue followed by a target. This task has been widely used to assess attentional mechanisms in children with Autism Spectrum Disorder (ASD) [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. This task has also been adapted for ADHD studies, as children with ADHD often demonstrate reduced responsiveness to social cues and frequently exhibit comorbid features with ASD [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan additionalcitationids=\"CR16 CR17\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Most previous studies have focused on behavioral indices, such as key responses, to evaluate whether participants responded to social cues. However, relying solely on key-press responses limits a comprehensive understanding of the attentional shifting process.\u003c/p\u003e\u003cp\u003eThe introduction of eye-tracking technology into gaze-cueing tasks has enabled more detailed assessments of attentional processes, including eye movements and fixation patterns [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. For instance, several studies have analyzed fixation durations within cue areas to examine gaze responsiveness [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Nevertheless, these studies primarily focus on cue utilization prior to target detection, rather than capturing the attentional shifting that occurs during early-stage orienting behavior in attentional shifts. While some studies have used various tasks, such as memory tasks, to differentiate ADHD, studies specifically targeting attention movement within gaze-cueing tasks remain limited [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAlthough most previous studies have focused on higher-level attentional shifting within complex tasks [\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], the present study is the first to reveal subtle delays in low-level attentional shifting in children with ADHD using a simplified gaze-cueing task with minimal complexity. Therefore, this study aims to investigate low-level attentional shifting characteristics in children with ADHD using a simplified gaze-cueing task with reduced distractors. By analyzing both gaze movements and behavioral response data during a straightforward goal-oriented task, we seek to determine whether children with ADHD exhibit slower orienting behavior compared to TD children.\u003c/p\u003e\u003cp\u003eThe objective of this study is to identify characteristic differences in attentional shifting between children with ADHD and TD children using a low-difficulty gaze-cueing task. Specifically, we aim to determine whether there are group differences in low-level attention shifts under conditions with minimal distractors, assessed through behavioral responses and eye-tracking data.\u003c/p\u003e\u003cp\u003eTo this end, the following hypotheses were formulated:\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eIn a simple gaze-cueing task, eye movement data will more effectively differentiate children with ADHD than key-press-based behavioral data.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eUnder minimal distractor interference, children with ADHD will exhibit less efficient orienting behavior compared to TD children.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eChildren with ADHD will show slower attentional shifting and increased reliance on peripheral vision rather than direct target detection.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1. Participants\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eA total of 44 TD children and 28 early elementary school-aged children with ADHD, aged 6\u0026ndash;9 years, were initially recruited between August 2021 and February 2022. TD children were recruited through an online community and local flyers, whereas children with ADHD were recruited from Korean medical centers and university hospitals. Eligibility was determined using an online application form that included exclusion criteria to ensure participants met the inclusion requirements.\u003c/p\u003e\u003cp\u003eParticipants in the ADHD group had recently been diagnosed by a psychiatrist and had been on ADHD medication for less than one year. Those exhibiting overt hyperactivity were excluded, as such behavior could lead to easy identification of ADHD status. Participants with visual acuity below 0.3 were excluded due to potential difficulties in perceiving visual cues during the task. Additional exclusion criteria included diagnoses of other developmental disorders, psychiatric illnesses, congenital genetic conditions, neurological diseases, or a history of acquired brain injury.\u003c/p\u003e\u003cp\u003eFurther exclusions were made during data processing. Five children diagnosed with both ASD and ADHD were excluded to eliminate potential confounds from comorbid developmental disorders. Additionally, participants with an intelligence quotient below 70, as assessed by the Wechsler Intelligence Scale for Children\u0026mdash;Fourth Edition, were excluded to control for borderline intellectual disability. To ensure clearer group distinctions, children in the ADHD group with Korean ADHD Rating Scale (KARS) inattention scores below 6 (n\u0026thinsp;=\u0026thinsp;1), and children in the TD group with KARS inattention scores above 10 (n\u0026thinsp;=\u0026thinsp;4), were excluded.\u003c/p\u003e\u003cp\u003eParticipants were excluded if they failed the eye-tracker calibration (n\u0026thinsp;=\u0026thinsp;9) or if their average visual deviation exceeded 2\u0026deg; during the validation trial (n\u0026thinsp;=\u0026thinsp;5), even if initial calibration was successful. Although no sessions were terminated due to participant difficulty, data from one child in each group were lost due to equipment failure and were excluded. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the overall flow chart.\u003c/p\u003e\u003cp\u003eThe final sample consisted of 45 children: 19 with ADHD and 26 TD children. This study was approved by the Central Research Facilities Research Ethics Board of the Ulsan National Institute of Science and Technology (UNISTIRB-20-62-A). All research procedures were performed in accordance with the relevant guidelines and regulations, including the Declaration of Helsinki. Written informed consent was obtained from the parents or legal guardians of all participants prior to their participation in the study.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2. Measurement\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eAdditional exclusion criteria were implemented through self-reported surveys and clinical diagnoses to ensure eligibility in both clinical and nonclinical groups. The KARS was used as a primary screening tool. The scale includes two subscales: inattention and hyperactivity/impulsivity. As the study focused on recruiting children with ADHD who exhibited minimal hyperactivity, inattention scores were used as the primary screening criterion. In a previous study, the average inattention score in the ADHD group was 11.22 (standard deviation [SD]\u0026thinsp;=\u0026thinsp;5.75). Therefore, ADHD participants with scores 1 SD below the mean (\u0026le;\u0026thinsp;5) were excluded. TD children were excluded if their inattention score exceeded 10 or their total KARS score surpassed the typical range for their age (\u0026ge;\u0026thinsp;18) [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. KARS scores were also used as predictor variables to assess group differences.\u003c/p\u003e\u003cp\u003eThe Childhood Autism Rating Scale (CARS) was used to screen for autism-related symptoms that might affect performance on the eye-tracking task (Cronbach\u0026rsquo;s alpha\u0026thinsp;=\u0026thinsp;0.87). CARS is a widely used screening tool applicable regardless of age or cognitive ability. Participants with a total score of 37 or higher were considered to be at high risk for autism and were excluded from the study [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. However, none of the participants exceeded this cutoff.\u003c/p\u003e\u003cp\u003eKovac\u0026rsquo;s Children\u0026rsquo;s Depression Inventory (CDI) was used to assess baseline levels of depression [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e], and the State-Trait Anxiety Inventory for Children (STAI-C) was used to measure anxiety [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. The CDI, adapted for children aged 8\u0026ndash;13, comprises 27 multiple-choice items rated 0 to 2. It comprises 27 multiple-choice items, each rated on a scale from 0 to 2. A total score of 22 or above indicates mild depression, 26 or above indicates moderate depression, and 29 or above indicates severe depression. The Cronbach\u0026rsquo;s alpha for the CDI in previous studies was 0.898 [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. The STAI-C measures anxiety in children using 20 items each for state and trait anxiety. In this study, we used a validated Korean version adapted for elementary school students. Each item is rated on a scale from 1 (not at all) to 3 (often), with higher scores indicating greater anxiety levels (Cronbach\u0026rsquo;s alpha\u0026thinsp;=\u0026thinsp;0.892) [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eBehavioral data were collected through keyboard inputs, measuring RT for each trial when the participant pressed a key corresponding to the target\u0026rsquo;s location and accuracy. Eye-tracking data were collected using a Tobii TX 300 eye tracker with a 300 Hz sampling rate. The monitor was 23\u0026rdquo;, with a screen resolution of 1920 \u0026times; 1080, and data were collected using the Titta toolbox supported by the Tobii Pro SDK in MATLAB [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Eye movement data were smoothed using the Savitzky\u0026ndash;Golay filter (sgolayfilt) with a polynomial order of 9 and a window size of 21. Eye movements with velocities\u0026thinsp;\u0026ge;\u0026thinsp;30\u0026deg;/s were classified as saccades, and those\u0026thinsp;\u0026lt;\u0026thinsp;30\u0026deg;/s were classified as fixations. The distance between the participant\u0026rsquo;s eyes and the monitor was maintained at 65 cm using a chin rest. The cue, target, and distractor were presented at the four corners of the screen, each positioned 0.2 x-units and 0.2-y units from the center, with a diameter of 77 pixels.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3. Material-gaze-cueing task\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eTo design a task related to joint attention (JA), we adapted Friesen and Kingstone\u0026rsquo;s gaze-cueing paradigm [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. In this task, social stimuli (faces) and nonsocial stimuli (arrows) were used to cue the target either congruently or incongruently (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). A cue appeared at the center of the screen and then shifted to point to one of the four corners for 500 ms. The target appeared following a randomized stimulus onset asynchrony (SOA) of 250, 500, or 750 ms. In the congruent condition, the cue pointed in the same direction as the target, whereas in the incongruent condition, the cue pointed to a different corner. The target was presented as a red dot, while a distractor (a blue dot) appeared simultaneously at one of the four corners to introduce interference. In the congruent condition, the distractor appeared in a location not aligned with the cue\u0026rsquo;s direction. In the incongruent condition, the distractor appeared in the direction of the cue to mislead the child\u0026rsquo;s attention. The target and distractor each subtended 2\u0026deg; of visual angle, with a consistent viewing distance of 65 cm from the screen.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eTo prevent directional bias, the direction of the cue and the locations of the target and distractor were randomized so that each combination appeared with equal probability. The distractor was randomly positioned in one of the three corners of the screen not occupied by the target. The entire task lasted approximately 15 min. There were four blocks, each consisting of 36 trials, and each block took about 9.2 min to complete. Social and nonsocial cues alternated between blocks, and the number of congruent and incongruent trials was evenly distributed within each block. A break of approximately 1 min was provided after the completion of the first two blocks.\u003c/p\u003e\u003cp\u003eBefore the start of blocks 1 and 3, a 6-point calibration task was performed to set up the eye tracker. Calibration was performed using the Tobii Pro Eye Tracker Manager software (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://developer.tobiipro.com/eyetrackermanager/etm-installation-information.html\u003c/span\u003e\u003cspan address=\"https://developer.tobiipro.com/eyetrackermanager/etm-installation-information.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Following calibration, a validation task was conducted to confirm calibration accuracy. A point was considered valid if its accuracy was within 2\u0026deg; of visual angle. Validation included three points; data were excluded if more than one point failed. Based on this criterion, 10 TD and 2 children with ADHD were excluded. The average validation accuracy among included participants was 1.84\u0026deg; for TD and 1.57\u0026deg; for ADHD children.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4. Data analysis\u003c/h2\u003e\u003cdiv id=\"Sec7\" class=\"Section3\"\u003e\u003ch2\u003e2.4.1. Definition of features\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThis study extracted both behavioral and eye-tracking features. Behavioral features focused on accuracy and RT, which are commonly used metrics in previous studies [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. RT was calculated for each stimulus, regardless of response correctness. Accuracy was calculated per condition, and RT metrics were derived using the median RT values from the trials within each condition. Additionally, RT variability, an indicator of sustained attention, was computed as the SD of RT values [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. The definitions and descriptions of eye-tracking features are provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e in Multimedia Appendix. JA and peripheral vision behaviors were assessed only in blocks involving social stimuli.\u003c/p\u003e\u003cp\u003eJA was defined using AOI centered on the cue (that is, the eye stimuli), and the surrounding screen was divided into quadrants, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. A gaze was classified as JA if it shifted from the central cue AOI to the quadrant indicated by the cue during the target detection period (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB-a).\u003c/p\u003e\u003cp\u003eIf the participant\u0026rsquo;s gaze remained fixed on the facial stimuli during the target detection period, without active gaze shifts toward the target location, this behavior was interpreted as target detection via \u0026ldquo;peripheral vision\u0026rdquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB-b).\u003c/p\u003e\u003cp\u003eTo reduce the impact of outliers, continuous variables (except those capturing behavior occurrence) were computed using median values rather than mean values. For instance, the RT metric for social cues in the congruent condition at an SOA of 0.5 s was calculated as the median RT value for that specific condition. For binary variables (0 or 1), e.g., JA presence or peripheral vision usage, mean values represented the proportion of trials exhibiting each behavior.\u003c/p\u003e\u003cp\u003eThe gaze-cueing effect (GCE) was computed by subtracting the values for congruent stimuli from those for incongruent stimuli under each SOA condition during the target detection phase, in trials with social stimuli [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. For example, GCE in RT at an SOA of 0.5 s was calculated by subtracting the RT under the congruent condition from that under the incongruent condition, both for social cues at that specific SOA.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section3\"\u003e\u003ch2\u003e2.4.2. Statistical analysis\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eTo assess the appropriateness of the participant pool and identify group-level differences, sociodemographic and clinical characteristic data were analyzed using the Mann\u0026ndash;Whitney U rank test, a nonparametric method used to compare the distributions of variables between ADHD and TD groups [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. This test was employed to identify features with statistically significant differences between the groups. To correct for multiple comparisons, the Benjamini\u0026mdash;Hochberg procedure was applied using the \u003cem\u003emultipletest\u003c/em\u003e function from the \u003cem\u003eStatsmodels\u003c/em\u003e library in Python [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Only features that remained significant after corrections were retained for further analysis.\u003c/p\u003e\u003cp\u003eNext, stepwise logistic regression was conducted for feature selection using the StepReg library in R Studio [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. This method employed bidirectional elimination, where variables were included based on their statistical significance. The entry and exit criteria for variable selection were based on Rao\u0026rsquo;s score test, with the significance threshold at 0.035. Both behavioral and eye-tracking features were subjected to this selection process, and the features identified as significant were used as inputs to train a combined predictive model.\u003c/p\u003e\u003cp\u003eLogistic regression modeling was then performed using selected features. To mitigate the risk of overfitting due to the small sample size, fivefold cross-validation was applied. Modeling was conducted in R Studio using the mlr3 package, and cross-validation was repeated 50 times to improve generalizability and reduce bias. During each iteration, label distribution was preserved across folds to maintain the balance between the ADHD and TD groups. For features found to be insignificant during modeling, feature importance was evaluated using the varImp function from the caret library in R Studio. Features with low importance scores were iteratively removed to determine whether this improved the performance of the model, particularly its area under the curve (AUC). This pruning process continued until no further AUC improvement was observed, resulting in the final feature set and model.\u003c/p\u003e\u003cp\u003eModel performance was evaluated based on the average accuracy, AUC, and F1 score (which incorporates both precision and recall) across the five validation folds. Consistency of the five cross-validation iterations was assessed using Cohen\u0026rsquo;s kappa coefficient, which measures the level of agreement between the actual labels and the predicted labels. Cohen\u0026rsquo;s kappa values were interpreted as follows:\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e0: Poor agreement\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e0\u0026ndash;0.2: Slight agreement\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e0.21\u0026ndash;0.40: Fair agreement\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e0.41\u0026ndash;0.60: Moderate agreement\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e0.61\u0026ndash;0.80: Substantial agreement\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e0.81\u0026ndash;1.00: Perfect agreement [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.1. Sociodemographic and clinical characteristics of participants\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe ADHD and TD groups were comparable in age, as both consisted of early elementary school\u0026ndash;aged children. Although there was a difference in sex distribution, this disparity was expected due to the higher prevalence of ADHD diagnoses in males. The KARS, which measures ADHD symptom severity, revealed significant group differences in both the inattention and hyperactivity/impulsivity subscales. Additionally, scores on the CARS differed significantly, supporting prior research indicating that children with ADHD tend to exhibit more autistic traits than their TD peers [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. No significant differences were found between the groups in state or trait anxiety. However, depression scores were significantly higher in the ADHD group (p\u0026thinsp;=\u0026thinsp;0.003). Detailed sociodemographic information is summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSociodemographic and clinical characteristics of each group.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMeasures (mean\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e\u003cem\u003eSD\u003c/em\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eADHD (n\u0026thinsp;=\u0026thinsp;19)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTD (n\u0026thinsp;=\u0026thinsp;27)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eP value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.7\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e1.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e7.9\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e0.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.709\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex (male n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e16 (84.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12 (44.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKARS total score\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e22.4\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e11.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:5.6\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e4.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInattention score\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e13.7\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e6.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.5\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e2.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHyperactivity/impulsivity score\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e8.7\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e6.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.1\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e2.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCARS\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e18.7\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e5.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e15.1\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e0.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCDI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e13.0\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e6.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e7.7\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e3.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.003\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSTAI-C\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTrait anxiety\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e30.3\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e5.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e28.3\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e5.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.259\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eState anxiety\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e33.3\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e6.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e30.6\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:4.9\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.176\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003e\u003cem\u003e3.2. Hypothesis 1: During a simple gaze-cueing task, eye movement data can distinguish children with ADHD more effectively than behavioral data\u003c/em\u003e\u003c/p\u003e\u003cp\u003eThe Mann\u0026ndash;Whitney U test was used to compare group differences across measured variables. As summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, trials using social cues showed significant group differences in the number of saccades, saccade velocity, and the SD of fixation locations. Approximately half of the significantly different features were related to JA.\u003c/p\u003e\u003cp\u003eThe frequency of JA was higher in the TD group compared to the ADHD group. Conversely, the difference in the number of saccades between incongruent and congruent conditions was larger in the ADHD group, with this contrast reaching high significance (p\u0026thinsp;\u0026lt;\u0026thinsp;.001).\u003c/p\u003e\u003cp\u003eIn trials utilizing nonsocial cues, significant indicators included the number of saccades, saccade velocity, SD of fixation locations, saccade length, and the ratio of null data (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eDistribution of eye-tracking features between ADHD and TD groups, selected using the Mann\u0026ndash;Whitney U test.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStimuli\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eData sort\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eFeature (SOA/congruency)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eADHD\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTD\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCohen\u0026rsquo;s D\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003ep-value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"10\" rowspan=\"11\"\u003e\u003cp\u003eSocial\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"9\" rowspan=\"10\"\u003e\u003cp\u003eEye\u003c/p\u003e\u003cp\u003emovement\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNumber of saccades\u003c/p\u003e\u003cp\u003e(0.25 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e6.89\u0026thinsp;\u0026plusmn;\u0026thinsp;6.03\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e10.3\u0026thinsp;\u0026plusmn;\u0026thinsp;5.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.594\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNumber of saccades\u003c/p\u003e\u003cp\u003e(0.5 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e5.89\u0026thinsp;\u0026plusmn;\u0026thinsp;4.33\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e10.46\u0026thinsp;\u0026plusmn;\u0026thinsp;6.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.037\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNumber of saccades\u003c/p\u003e\u003cp\u003e(0.75 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e6.66\u0026thinsp;\u0026plusmn;\u0026thinsp;5.57\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e9.69\u0026thinsp;\u0026plusmn;\u0026thinsp;5.49\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.548\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSaccade velocity\u003c/p\u003e\u003cp\u003e(0.5 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e89.56\u0026thinsp;\u0026plusmn;\u0026thinsp;22.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e100.09\u0026thinsp;\u0026plusmn;\u0026thinsp;24.45\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.447\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSaccade velocity\u003c/p\u003e\u003cp\u003e(0.75 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e89.78\u0026thinsp;\u0026plusmn;\u0026thinsp;21.21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e100.08\u0026thinsp;\u0026plusmn;\u0026thinsp;25.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.428\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.045\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSD of fixation locations\u003c/p\u003e\u003cp\u003e(0.25 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e4.91\u0026thinsp;\u0026plusmn;\u0026thinsp;3.16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e6.29\u0026thinsp;\u0026plusmn;\u0026thinsp;3.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.443\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.045\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRate of JA\u003c/p\u003e\u003cp\u003e(0.25 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.24\u0026thinsp;\u0026plusmn;\u0026thinsp;0.14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.34\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.804\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRate of JA\u003c/p\u003e\u003cp\u003e(0.25 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.2\u0026thinsp;\u0026plusmn;\u0026thinsp;0.13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.3\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.834\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.037\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eGCE in the saccade number\u003c/p\u003e\u003cp\u003e(0.5 s/incong\u0026ndash;cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e3.13\u0026thinsp;\u0026plusmn;\u0026thinsp;2.99\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e-0.57\u0026thinsp;\u0026plusmn;\u0026thinsp;3.02\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.233\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.009\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRate of peripheral vision (0.25 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.64\u0026thinsp;\u0026plusmn;\u0026thinsp;0.16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.56\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.627\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.046\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBehavior\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eGCE in RT\u003c/p\u003e\u003cp\u003e(0.5 s/incong\u0026ndash;cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.12\u0026thinsp;\u0026plusmn;\u0026thinsp;0.15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.03\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.855\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.045\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"13\" rowspan=\"14\"\u003e\u003cp\u003eNon-\u003c/p\u003e\u003cp\u003esocial\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"12\" rowspan=\"13\"\u003e\u003cp\u003eEye\u003c/p\u003e\u003cp\u003emovement\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSaccade length\u003c/p\u003e\u003cp\u003e(0.75 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e11.3\u0026thinsp;\u0026plusmn;\u0026thinsp;6.02\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e16.59\u0026thinsp;\u0026plusmn;\u0026thinsp;8.32\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.708\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.041\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSaccade length\u003c/p\u003e\u003cp\u003e(0.75 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e20.12\u0026thinsp;\u0026plusmn;\u0026thinsp;13.15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e27.02\u0026thinsp;\u0026plusmn;\u0026thinsp;10.67\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.588\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNumber of saccades\u003c/p\u003e\u003cp\u003e(0.25 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e5.71\u0026thinsp;\u0026plusmn;\u0026thinsp;4.53\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e9.22\u0026thinsp;\u0026plusmn;\u0026thinsp;5.5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.685\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.037\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNumber of saccades\u003c/p\u003e\u003cp\u003e(0.25 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e6.61\u0026thinsp;\u0026plusmn;\u0026thinsp;4.72\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e10.81\u0026thinsp;\u0026plusmn;\u0026thinsp;6.28\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.739\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.037\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSaccade velocity\u003c/p\u003e\u003cp\u003e(0.5 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e96.97\u0026thinsp;\u0026plusmn;\u0026thinsp;23.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e105.08\u0026thinsp;\u0026plusmn;\u0026thinsp;21.53\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.366\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSaccade velocity\u003c/p\u003e\u003cp\u003e(0.75 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e83.94\u0026thinsp;\u0026plusmn;\u0026thinsp;18.39\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e94.76\u0026thinsp;\u0026plusmn;\u0026thinsp;19.11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.575\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSaccade velocity\u003c/p\u003e\u003cp\u003e(0.75 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e93.67\u0026thinsp;\u0026plusmn;\u0026thinsp;20.15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e104.71\u0026thinsp;\u0026plusmn;\u0026thinsp;23.69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.495\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.037\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSD of fixation locations\u003c/p\u003e\u003cp\u003e(0.25 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e7.13\u0026thinsp;\u0026plusmn;\u0026thinsp;3.46\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e9.08\u0026thinsp;\u0026plusmn;\u0026thinsp;3.11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.596\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSD of fixation locations\u003c/p\u003e\u003cp\u003e(0.5 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e7.56\u0026thinsp;\u0026plusmn;\u0026thinsp;3.43\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e9.43\u0026thinsp;\u0026plusmn;\u0026thinsp;3.3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.557\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSD of fixation locations\u003c/p\u003e\u003cp\u003e(0.75 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e6.39\u0026thinsp;\u0026plusmn;\u0026thinsp;3.76\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e9.3\u0026thinsp;\u0026plusmn;\u0026thinsp;3.51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.805\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.037\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRate of null data\u003c/p\u003e\u003cp\u003e(0.5 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.07\u0026thinsp;\u0026plusmn;\u0026thinsp;0.13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.0\u0026thinsp;\u0026plusmn;\u0026thinsp;0.02\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.841\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.041\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRate of null data\u003c/p\u003e\u003cp\u003e(0.75 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.08\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.01\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.808\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRate of null data\u003c/p\u003e\u003cp\u003e(0.75 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.05\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.0\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.801\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBehavior\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAccuracy (0.25 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.76\u0026thinsp;\u0026plusmn;\u0026thinsp;0.35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.98\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.981\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.037\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eEye\u003c/p\u003e\u003cp\u003emovement\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRate of null data\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.11\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.05\u0026thinsp;\u0026plusmn;\u0026thinsp;0.04\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.957\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.041\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSD of null rate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.15\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.727\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBehavior\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSD of RT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e\u003cp\u003e0.55\u0026thinsp;\u0026plusmn;\u0026thinsp;0.52\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e\u003cp\u003e0.26\u0026thinsp;\u0026plusmn;\u0026thinsp;0.17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.818\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eOverall, the TD group exhibited more frequent and faster eye movements during the target detection period, along with greater variability in gaze behavior.\u003c/p\u003e\u003cp\u003eThree behavioral features were significantly different between the groups. Accuracy under the nonsocial cue condition at an SOA of 0.25 s during incongruent trials differed significantly (p\u0026thinsp;=\u0026thinsp;0.008), whereas accuracy under other conditions did not. The SD of RT differed significantly across all conditions (p\u0026thinsp;=\u0026thinsp;0.033), and the GCE in RT at an SOA of 0.5 s also showed a significant group difference (p\u0026thinsp;=\u0026thinsp;0.044).\u003c/p\u003e\u003cp\u003eStepwise regression analysis identified accuracy and GCE in RT as significant predictors in the behavioral model. For the eye movement model, key features included the number of saccades, GCE in number of saccades, JA rate, and the proportion of null data. Notably, approximately half of the selected features in both models were related to JA.\u003c/p\u003e\u003cp\u003eIrrespective of the data source, the model included the SD of the RT, number of saccades, GCE in the number of saccades, and SD of fixation locations when considering the entire dataset.\u003c/p\u003e\u003cp\u003eModel performance metrics derived from logistic regression using these features are presented in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eFeatures used for logistic regression and model performance by each dataset.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDataset\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFeatures used for the model\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eKappa\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAcc\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eF1 score\u003c/p\u003e\u003cp\u003e(Precision/Recall)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAUC\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBehavior\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e- Accuracy (0.25s/incong)\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e- GCE in RT (0.5)\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e.435\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e.741\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e.788 (.746/.871)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e.828\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEye movement\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e- Number of saccade (0.25s/incong)\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e- GCE in number of saccade (0.5s)\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e- Rate of JA (0.25s/cong)\u003c/p\u003e\u003cp\u003e- Rate of null data\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e.644\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e.837\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e.853 (.863/.872)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e.929\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e- SD of RT\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e- Number of saccade (0.25s/cong)\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e- GCE in number of saccade (0.5s)\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e- SD of fixation locations (0.25s/incong)\u003csup\u003e\u0026middot;\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e.716\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e.870\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e.882 (.883/.901)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e.911\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003csup\u003e*\u003c/sup\u003e indicates significant variables in the logistic regression (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05)\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe eye movement model outperformed the behavioral model in both accuracy (0.837 vs. 0.741) and AUC (0.929 vs. 0.828). While the accuracy and AUC of the eye movement model were comparable to those of the combined model, the kappa statistic was higher for the integrated model that incorporated both behavioral and eye-tracking data (kappa\u0026thinsp;=\u0026thinsp;0.716). The behavior-only model demonstrated a moderate level of agreement (k\u0026thinsp;=\u0026thinsp;0.435), whereas both the eye movement-only and combined models achieved significant agreement according to established kappa interpretation guidelines [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.3. Hypothesis 2 : Under minimal distractor interference, children with ADHD show\u003c/h2\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003e\u003cem\u003eless efficient goal-oriented attentional shifting compared to TD children\u003c/em\u003e\u003c/p\u003e\u003cp\u003eAlthough RT did not differ significantly between the groups (p\u0026thinsp;=\u0026thinsp;0.707), a trend emerged suggesting that the TD group produced a higher number of saccades (p\u0026thinsp;=\u0026thinsp;0.057, d= -0.516). This trend prompted an examination of fixation duration per saccade, calculated by dividing total fixation duration by the number of saccades during target detection. Results showed that children with ADHD had significantly longer fixation duration per saccade compared to TD children (p\u0026thinsp;=\u0026thinsp;0.017, d\u0026thinsp;=\u0026thinsp;0.799), indicating less efficient gaze shifting. However, accuracy was lower in the ADHD group compared to the TD group (p\u0026thinsp;=\u0026thinsp;0.086). Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrates the distributions of these features.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eFollowing the inclusion of eye movement metrics in the subset selection process, model performance remained consistent regardless of whether features were selected from eye-tracking data alone or in combination with behavioral data. The key features identified included:\u003c/p\u003e\u003cp\u003e(1) GCE in number of saccades at SOA\u0026thinsp;=\u0026thinsp;0.5 s, (2) rate of null data under nonsocial cues (SOA\u0026thinsp;=\u0026thinsp;0.5 s, incongruent), and (3) fixation duration per saccade in social cue trials. Notably, the \u0026ldquo;fixation duration per saccade\u0026rdquo; was a novel metric identified specifically under social cue conditions during the target detection period. Incorporating this metric improved the performance of the logistic regression model. The updated model achieved an AUC of 0.940 and an accuracy of 0.860, demonstrating a significant improvement over the previous version (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eComparison of the model with and without duration in each fixation.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e\u003cp\u003eDuration in each fixation -\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e\u003cp\u003eDuration in each fixation +\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{\\beta\\:}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003eZ\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e(p-value)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{\\beta\\:}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003eZ\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e(p-value)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e\u003cp\u003eFeatures\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSD of RT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e14.771\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.467 (.014)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNumber of saccade\u003c/p\u003e\u003cp\u003e(0.25 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026minus;\u0026thinsp;.842\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-2.348 (.019)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSD of fixation locations\u003c/p\u003e\u003cp\u003e(0.25 s/incong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026minus;\u0026thinsp;.666\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-1.676 (.094)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGCE in number of saccade (0.5 s)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.901\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.177 (.030)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u003cp\u003e1.306\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e2.485 (.013)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003enull_rate in nonsocial\u003c/p\u003e\u003cp\u003e(0.5 s/cong)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u003cp\u003e114.266\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e2.449 (.014)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003efixation duration per saccade\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u003cp\u003e43.157\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e2.646 (.008)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003eKappa\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e\u003cp\u003e.716\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e0.696\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003eAcc\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e\u003cp\u003e.870\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e.860\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003eF1 score\u003c/p\u003e\u003cp\u003e(precision/recall)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e\u003cp\u003e.882 (.883/.901)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e0.876 (0.879/0.891)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAUC\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e\u003cp\u003e\u003cb\u003e.911\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e\u003cb\u003e0.940\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c8\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThe correlations between selected model features and the ADHD symptom severity, as measured by the KARS, were evaluated (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e in Multimedia Appendix). Fixation duration per saccade exhibited the strongest correlation with inattention (r\u0026thinsp;=\u0026thinsp;0.46, p\u0026thinsp;\u0026lt;\u0026thinsp;.01) and hyperactivity scores (r\u0026thinsp;=\u0026thinsp;0.36, p\u0026thinsp;\u0026lt;\u0026thinsp;.01). Features such as GCE in number of saccades at an SOA of 0.5 s and the rate of null data in nonsocial cues at an SOA of 0.5 s under incongruent conditions showed a significant correlation with the inattention KARS scores (p\u0026thinsp;\u0026lt;\u0026thinsp;.01). Conversely, among other features, only the SD of RT showed a significant correlation with the hyperactivity KARS scores (r\u0026thinsp;=\u0026thinsp;0.32, p\u0026thinsp;\u0026lt;\u0026thinsp;.01), excluding the duration at each fixation point. Other features included in the model did not show statistically significant correlations with KARS scores (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e in Multimedia Appendix).\u003c/p\u003e\u003cp\u003e\u003cem\u003e3.4. Hypothesis 3: Children with ADHD exhibit slower attentional shifting and rely more on peripheral vision rather than actively detecting the target.\u003c/em\u003e\u003c/p\u003e\u003cp\u003eIn trials involving social cues, the rate of JA was significantly higher in the TD group than in the ADHD group (p\u0026thinsp;=\u0026thinsp;0.042, d = -0.827), regardless of SOA or cue congruency. Furthermore, despite comparable RTs, the ADHD group demonstrated significantly longer fixation durations per saccade than the TD children (p\u0026thinsp;=\u0026thinsp;0.017, d\u0026thinsp;=\u0026thinsp;0.799). Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows representative eye movement patterns for the two groups.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThese findings suggest that children with ADHD tend to restrict their gaze to the AOI surrounding the central cue, rather than executing goal-directed gaze shifts.\u003c/p\u003e\u003cp\u003eAdditionally, the ADHD group exhibited a significantly higher frequency of peripheral vision use during the target detection period compared to the TD group, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e (p\u0026thinsp;=\u0026thinsp;0.046, d\u0026thinsp;=\u0026thinsp;0.481).\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThis study utilized a gaze-cuing task to collect both behavioral and eye movement data from children, with the aim of identifying indicators of ADHD and developing an early-stage classification model. Our findings supported all three hypotheses and provided new insights into low level of attentional and executive functioning deficits in children with ADHD, even in low-demand tasks.\u003c/p\u003e\u003cp\u003eFirst, consistent with our initial hypothesis, this study highlights the diagnostic utility of eye movement data in detecting ADHD, particularly in low-demand tasks where behavioral differences are subtle. Unlike previous studies that reported significant differences in accuracy among children with ADHD [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e], we observed significant behavioral differences in accuracy only under the social cue condition with a SOA of 0.25 s and incongruent cues. Conversely, the model developed in this study achieved relatively high performance using only eye movement indicators (AUC\u0026thinsp;=\u0026thinsp;0.929, accuracy\u0026thinsp;=\u0026thinsp;0.837). These results suggest that the gaze-cueing task employed in this study may not have been sufficiently demanding to differentiate the groups based on behavioral indicators alone [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eSecond, our findings support the hypothesis that children with ADHD exhibit less efficient goal-oriented attentional shifting, even in environments with minimal distractor interference. This was demonstrated by prolonged fixation durations and reduced saccadic frequencies during the target detection behavior. These eye movement patterns suggest delayed attentional disengagement from irrelevant areas and reduced engagement in active visual search\u0026mdash;both of which are essential for effective attentional shifting. These findings align with prior research linking reduced saccadic frequency to impaired attentional control and suggesting that limited exploratory eye movements reflect executive dysfunction in ADHD [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eMoreover, fixation duration per saccade during target detection emerged as the most predictive feature in our classification model, showing strong correlations with KARS inattention (r\u0026thinsp;=\u0026thinsp;0.46, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01) and hyperactivity scores (r\u0026thinsp;=\u0026thinsp;0.36, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). Prior meta-analyses have indicated that small to moderate deficits in attentional shifting are associated with inattention and hyperactivity [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Our findings further suggest that prolonged fixations reflect more than motor slowing\u0026mdash;they may indicate a reduced capacity for flexible attentional engagement.\u003c/p\u003e\u003cp\u003eWhile previous work has shown that ADHD may not impair set-shifting speed when high-level executive functions remain intact [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e], our findings suggest deficits at more automatic, early stages of orienting. This supports a hierarchical model of attentional control, where ADHD symptoms manifest at multiple cognitive levels. The subtle nature of these low-level impairments highlights the importance of examining nontraditional indicators, such as prolonged fixations, to capture executive dysfunction beyond what standard neuropsychological tasks can detect.\u003c/p\u003e\u003cp\u003eThird, our data support the hypothesis that children with ADHD rely more on peripheral vision than on active, intentional target detection. Compared with TD children, the ADHD group exhibited both a lower rate of JA and a higher rate of peripheral vision use. JA serves as a regulatory mechanism that facilitates shared attention and social coordination [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. A reduced rate of JA in the ADHD group suggests impaired use of external cues for attentional guidance. This deficit may indicate overlapping impairments in social cue responsiveness, consistent with the high comorbidity between ADHD and ASD [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Given that reduced JA is a prominent early marker for ASD, our finding of diminished JA in children with ADHD supports the notion of shared difficulties in processing socially relevant cues [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAlthough a previous study reported increased face fixation time in children with ADHD across the task duration [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], this study adopts a more focused approach by examining eye movements immediately following gaze cue onset. This allowed us to assess the visual strategies employed specifically during goal-directed target detection. Our findings reveal that children with ADHD are more likely to rely on peripheral visual input and to engage in less structured visual scanning.\u003c/p\u003e\u003cp\u003eNotably, the ADHD group demonstrated a greater difference in saccade frequency between congruent and incongruent trials than the TD group. This pattern may reflect heightened sensitivity to conflicting cues, indicating reduced cognitive control in situations that demand attentional flexibility [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]\u003c/p\u003e\u003cp\u003eAcross all three hypotheses, eye movement patterns in the ADHD group consistently indicated delayed and inefficient attentional shifting, as well as difficulty using cues to guide attention. These impairments\u0026mdash;evident even in a distraction-free, low-demand task\u0026mdash;highlight fundamental executive dysfunctions in ADHD patients and suggest that deficits in attentional shifting impairments may serve as robust behavioral markers for early identification.\u003c/p\u003e\u003cp\u003eThis study presents several notable strengths. First, it employed a simple gaze-cueing task devoid of complex distractors to investigate low-level attentional shifting in children with ADHD. Although such paradigms are commonly used in developmental research, they have rarely been applied to detect subtle attention deficits in ADHD populations. Despite minimal task demands, significant group differences emerged in eye-tracking measures.\u003c/p\u003e\u003cp\u003eSecond, the study identified prolonged fixation duration during target detection as a novel and highly predictive marker of ADHD. This feature was linked to reduced attentional shifting efficiency and showed strong correlations with clinical indices of inattention and hyperactivity. Its inclusion in the classification model significantly improved predictive accuracy, underscoring its potential as a core feature for early ADHD detection.\u003c/p\u003e\u003cp\u003eFurthermore, by analyzing eye movement patterns, this study provided insight into the behavioral mechanisms underlying attentional shifts in ADHD. These findings revealed physiological indicators that may facilitate early identification, and could be integrated with other bio-signals\u0026mdash;such as electroencephalography\u0026mdash;to enhance diagnostic precision in future multimodal assessments.\u003c/p\u003e\u003cp\u003eLastly, despite the limited sample size, the study achieved high classification accuracy and AUC values using a minimal set of eye movement features. The use of cross-validation techniques helped reduce overfitting, suggesting these indicators may scale well in larger diagnostic frameworks.\u003c/p\u003e\u003cp\u003eNevertheless, some limitations should be acknowledged. The relatively low task complexity may have reduced behavioral differentiation between groups, although eye movement data effectively compensated for this limitation. Additionally, the small sample size may restrict the generalizability of the findings. Future studies with larger and more diverse samples are needed to validate and refine these findings.\u003c/p\u003e\u003cp\u003eThis study demonstrates that even a simple gaze-cueing task, analyzed using a logistic regression model based on behavioral and eye movement data, can objectively detect atypical attentional shifting in children with ADHD. Notably, eye-tracking indicators, particularly prolonged fixation durations per saccade and reduced saccade frequency, proved more sensitive than traditional behavioral responses in distinguishing ADHD, emerging as strong markers linked to core symptoms. In particular, fixation duration during target detection showed significant correlations with inattention and hyperactivity.\u003c/p\u003e\u003cp\u003eThese findings underscore the potential of combining simplified gaze-cueing tasks with eye-tracking technology to reveal subtle executive dysfunctions and enable early, efficient ADHD screening. The strong classification performance further supports the utility of these markers. Future research should expand sample sizes and explore these markers across diverse populations and settings to enhance generalizability and refine diagnostic tools.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Contributions:\u003c/strong\u003e L.S.M. conceived the study, collected data, performed preprocessing and statistical analyses, and wrote the original draft. L.S.I. contributed to the conceptualization, experimental design, discussion, and revision of the manuscript. J.I.J. collected data and contributed to the original draft. J.J.H. collected data and contributed to data preprocessing and statistical analyses. P.H.J. contributed to the conceptualization. K.M.K. contributed to the conceptualization, provided advice on data analysis, revised the manuscript, and acquired funding. Z.T. contributed to the discussion and revision of the manuscript. S.S. contributed to the conceptualization and funding acquisition. J.D.Y. contributed to the conceptualization and methodology, revised the manuscript, supervised the project, and acquired funding. All authors reviewed and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInstitutional Review Board Statement:\u0026nbsp;\u003c/strong\u003eThis study was approved by the Central Research Facilities Research Ethics Board of the Ulsan National Institute of Science and Technology (UNISTIRB-20-62-A).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed Consent Statement:\u0026nbsp;\u003c/strong\u003eInformed consent was obtained from all subjects involved in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statement:\u003c/strong\u003e The datasets generated and/or analyzed during this study are not publicly available due to participant confidentiality agreements, but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2020M3E5D9080787).\u0026nbsp;This research was supported by a grant of the Korea Health Technology R\u0026amp;D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health \u0026amp; Welfare, Republic of Korea (Grant Number: HI22C0646).\u0026nbsp;This study was supported by the U-K (UNIST-Korea) research brand program (1.230016.01) funded by UNIST (Ulsan National Institute of Science \u0026amp; Technology).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of Interest:\u003c/strong\u003e The authors have no conflict of interest to declare.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBiederman, J. Attention-deficit/hyperactivity disorder: a selective overview. \u003cem\u003eBiol. Psychiatry\u003c/em\u003e. \u003cb\u003e57\u003c/b\u003e, 1215\u0026ndash;1220 (2005).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFaraone, S. V., Biederman, J. \u0026amp; Mick, E. The age-dependent decline of attention deficit hyperactivity disorder: a meta-analysis of follow-up studies. \u003cem\u003ePsychol. Med.\u003c/em\u003e \u003cb\u003e36\u003c/b\u003e, 159\u0026ndash;165 (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePennington, B. F. \u0026amp; Ozonoff, S. Executive functions and developmental psychopathology. \u003cem\u003eJ. Child. Psychol. Psychiatry\u003c/em\u003e. \u003cb\u003e37\u003c/b\u003e, 51\u0026ndash;87 (1996).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBrown, S. W. Timing and executive function: Bidirectional interference between concurrent temporal production and randomization tasks. \u003cem\u003eMem. Cognit\u003c/em\u003e. \u003cb\u003e34\u003c/b\u003e, 1464\u0026ndash;1471 (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBarkley, R. A. Behavioral inhibition, sustained attention, and executive functions: constructing a unifying theory of ADHD. \u003cem\u003ePsychol. Bull.\u003c/em\u003e \u003cb\u003e121\u003c/b\u003e, 65 (1997).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQian, Y. et al. The developmental trajectories of executive function of children and adolescents with attention deficit hyperactivity disorder. \u003cem\u003eRes. Dev. Disabil.\u003c/em\u003e \u003cb\u003e34\u003c/b\u003e, 1434\u0026ndash;1445 (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e\u0026Ccedil;ak, H. T. et al. The behavior rating inventory of executive function and continuous performance test in preschoolers with attention deficit hyperactivity disorder. \u003cem\u003ePsychiatry Investig\u003c/em\u003e. \u003cb\u003e14\u003c/b\u003e, 260\u0026ndash;265 (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShalev, L. et al. Conjunctive Continuous Performance Task (CCPT)\u0026mdash;A pure measure of sustained attention. \u003cem\u003eNeuropsychologia\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e, 2584\u0026ndash;2591 (2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBoshomane, T. T., Pillay, B. J. \u0026amp; Meyer, A. Mental flexibility (set-shifting) deficits in children with ADHD: a replication and extension study. \u003cem\u003eJ. Psychol. Afr.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 344\u0026ndash;349 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuizenga, H. M. et al. Task complexity enhances response inhibition deficits in childhood and adolescent attention-deficit/hyperactivity disorder: a meta-regression analysis. \u003cem\u003eBiol. Psychiatry\u003c/em\u003e. \u003cb\u003e65\u003c/b\u003e, 39\u0026ndash;45 (2009).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePosner, M. I. \u0026amp; Petersen, S. E. The attention system of the human brain. \u003cem\u003eAnnu. Rev. Neurosci.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 25\u0026ndash;42 (1990).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSherigar, S. S., Gamsa, A. H. \u0026amp; Srinivasan, K. Oculomotor deficits in attention deficit hyperactivity disorder: a systematic review and meta-analysis. \u003cem\u003eEye\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e, 1975\u0026ndash;1981 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSeernani, D. et al. Social and non-social gaze cueing in autism spectrum disorder, attention-deficit/hyperactivity disorder and a comorbid group. \u003cem\u003eBiol. Psychol.\u003c/em\u003e \u003cb\u003e162\u003c/b\u003e, 108096 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStallworthy, I. C. et al. Variability in responding to joint attention cues in the first year is associated with autism outcome. \u003cem\u003eJ. Am. Acad. Child. Adolesc. Psychiatry\u003c/em\u003e. \u003cb\u003e61\u003c/b\u003e, 413\u0026ndash;422 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTemelt\u0026uuml;rk, R. D. et al. Dynamic eye-tracking evaluation of responding joint attention abilities and face scanning patterns in children with attention deficit hyperactivity disorder. \u003cem\u003eDev. Psychopathol.\u003c/em\u003e \u003cb\u003e36\u003c/b\u003e, 1190\u0026ndash;1201 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMarotta, A. et al. Impaired reflexive orienting to social cues in attention deficit hyperactivity disorder. \u003cem\u003eEur. Child. Adolesc. Psychiatry\u003c/em\u003e. \u003cb\u003e23\u003c/b\u003e, 649\u0026ndash;657 (2014).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTemelt\u0026uuml;rk, R. D. et al. Dynamic eye-tracking evaluation of responding joint attention abilities and face scanning patterns in children with attention deficit hyperactivity disorder. \u003cem\u003eDev. Psychopathol.\u003c/em\u003e \u003cb\u003e36\u003c/b\u003e, 1\u0026ndash;12 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYoo, J. H. et al. Development of an innovative approach using portable eye tracking to assist ADHD screening: a machine learning study. \u003cem\u003eFront. Psychiatry\u003c/em\u003e. \u003cb\u003e15\u003c/b\u003e, 1337595 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJang, S. J., Suh, D. S. \u0026amp; Byun, H. J. Normative study of the K-ARS (Korean ADHD Rating Scale) for parents. \u003cem\u003eJ. Korean Acad. Child. Adolesc. Psychiatry\u003c/em\u003e. \u003cb\u003e18\u003c/b\u003e, 38\u0026ndash;48 (2007).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee, N., Hong, Y. \u0026amp; Kim, Y. Screening instruments for autism spectrum disorder: mini review. \u003cem\u003eJ. Korean Neuropsychiatr Assoc.\u003c/em\u003e \u003cb\u003e58\u003c/b\u003e, 192\u0026ndash;201 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCho, S. \u0026amp; Lee, Y. Development of the Korean form of the Kovacs\u0026rsquo; Children\u0026rsquo;s Depression Inventory. \u003cem\u003eJ. Korean Neuropsychiatr Assoc.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e, 943\u0026ndash;956 (1990).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChoi, J. \u0026amp; Cho, S. Assessment of anxiety in children: Reliability and validity of revised children\u0026rsquo;s manifest anxiety scale. \u003cem\u003eJ. Korean Neuropsychiatr Assoc.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e, 691\u0026ndash;702 (1990).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee, W. J. et al. The propensity to depression and anxiety in children. \u003cem\u003eJ. Korean Acad. Fam Med.\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, 828\u0026ndash;837 (1998).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNiehorster, D. C., Andersson, R. \u0026amp; Nystr\u0026ouml;m, M. Titta: a toolbox for creating PsychToolbox and Psychopy experiments with Tobii eye trackers. \u003cem\u003eBehav. Res. Methods\u003c/em\u003e. \u003cb\u003e52\u003c/b\u003e, 1970\u0026ndash;1979 (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFriesen, C. K. \u0026amp; Kingstone, A. The eyes have it! Reflexive orienting is triggered by nonpredictive gaze. \u003cem\u003ePsychon Bull. Rev.\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e, 490\u0026ndash;495 (1998).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShin, W. G. et al. Individual differences in gaze-cuing effect are associated with facial emotion recognition and social conformity. \u003cem\u003eFront. Psychol.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 1219488 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMacFarland, T. W. \u0026amp; Yates, J. M. \u003cem\u003eIntroduction to Nonparametric Statistics for the Biological Sciences Using R\u003c/em\u003e (Springer, 2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBenjamini, Y. \u0026amp; Hochberg, Y. On the adaptive control of the false discovery rate in multiple testing with independent statistics. \u003cem\u003eJ. Educ. Behav. Stat.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 60\u0026ndash;83 (2000).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, J. et al. \u003cem\u003ePackage \u0026lsquo;StepReg\u0026rsquo;\u003c/em\u003e. (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWardhani, N. W. S. et al. IEEE,. Cross-validation metrics for evaluating classification performance on imbalanced data. In \u003cem\u003eProc. 2019 Int. Conf. Comput. Control Inform. Appl. (IC3INA)\u003c/em\u003e 45\u0026ndash;50 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHattori, J. et al. Are pervasive developmental disorders and attention-deficit/hyperactivity disorder distinct disorders? \u003cem\u003eBrain Dev.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 371\u0026ndash;374 (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJones, D. R. et al. Prevalence, severity, and co-occurrence of chronic physical health problems of persons with serious mental illness. \u003cem\u003ePsychiatr Serv.\u003c/em\u003e \u003cb\u003e55\u003c/b\u003e, 1250\u0026ndash;1257 (2004).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLambek, R. et al. Executive dysfunction in school-age children with ADHD. \u003cem\u003eJ. Atten. Disord\u003c/em\u003e. \u003cb\u003e15\u003c/b\u003e, 646\u0026ndash;655 (2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWillcutt, E. G. et al. Neuropsychological analyses of comorbidity between reading disability and attention deficit hyperactivity disorder: In search of the common deficit. \u003cem\u003eDev. Neuropsychol.\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e, 35\u0026ndash;78 (2005).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIrwin, L. N. et al. Do children with attention-deficit/hyperactivity disorder (ADHD) have set shifting deficits? \u003cem\u003eNeuropsychology\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e, 470\u0026ndash;481 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDrigas, A. \u0026amp; Karyotaki, M. Attentional control and other executive functions. \u003cem\u003eInt. J. Emerg. Technol. Learn.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 219\u0026ndash;230 (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMundy, P. et al. Brief report: joint attention and information processing in children with higher functioning autism spectrum disorders. \u003cem\u003eJ. Autism Dev. Disord\u003c/em\u003e. \u003cb\u003e46\u003c/b\u003e, 2555\u0026ndash;2560 (2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHours, C., Recasens, C. \u0026amp; Baleyte, J. M. ASD and ADHD comorbidity: What are we talking about? \u003cem\u003eFront. Psychiatry\u003c/em\u003e. \u003cb\u003e13\u003c/b\u003e, 837424 (2022).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"early-stage orienting behavior, ADHD, eye movement, attention shifts, classification, gaze-cueing task","lastPublishedDoi":"10.21203/rs.3.rs-7191610/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7191610/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eExploring early-stage orienting behavior is essential for elucidating the behavioral mechanisms underlying attentional shifts in attention deficit hyperactivity disorder (ADHD). However, traditional tasks lacking eye-tracking data often obscure these mechanisms. This study investigates low-level attentional shifting in ADHD using a simplified gaze-cueing task and explores classification markers via eye movement. Eye-tracking data were collected from 44 typically developing children and 28 children diagnosed with ADHD. We constructed a logistic regression model for classification purposes. Eye movement data alone yielded an accuracy of 0.84, comparable to the accuracy achieved using combined eye-tracking and behavioral data (0.87), underscoring the sensitivity of gaze-based features. Children with ADHD exhibited significantly prolonged fixation (p\u0026thinsp;=\u0026thinsp;.02, d\u0026thinsp;=\u0026thinsp;0.80) and marginally reduced saccade frequency (p\u0026thinsp;=\u0026thinsp;.06, d = \u0026minus;\u0026thinsp;0.52) during target detection, indicating delayed attentional shifting and diminished goal-directed attention. Prolonged fixation during target detection behavior emerged as the strongest predictor, correlating with both inattention and hyperactivity (r\u0026thinsp;=\u0026thinsp;.46; r\u0026thinsp;=\u0026thinsp;.36; both p\u0026thinsp;\u0026lt;\u0026thinsp;.01). Additionally, children with ADHD demonstrated lower joint attention and a greater reliance on peripheral vision. These findings highlight distinct gaze patterns under low cognitive load, revealing subtle mechanisms of executive dysfunction and potential early classification markers.\u003c/p\u003e","manuscriptTitle":"Exploring early-stage orienting behavior using an eye tracker for attention deficit hyperactivity disorder classification","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-01 09:26:25","doi":"10.21203/rs.3.rs-7191610/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-11-12T15:01:09+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-08T14:14:20+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-19T10:59:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"29907817978764616362187976681468490033","date":"2025-09-17T08:06:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"156651655949435217766652449625425907089","date":"2025-09-12T06:27:11+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-11T10:27:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"235523768212893563306743873721231332131","date":"2025-09-01T07:53:03+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-20T06:47:15+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-07-29T11:30:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-29T02:54:47+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-26T19:25:50+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-07-26T16:47:40+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5eba2d0e-f710-4d7d-9a0d-75d502dd7983","owner":[],"postedDate":"September 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":53614867,"name":"Health sciences/Diseases"},{"id":53614868,"name":"Biological sciences/Neuroscience"},{"id":53614869,"name":"Biological sciences/Psychology"},{"id":53614870,"name":"Social science/Psychology"}],"tags":[],"updatedAt":"2026-03-02T16:01:43+00:00","versionOfRecord":{"articleIdentity":"rs-7191610","link":"https://doi.org/10.1038/s41598-026-41419-0","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2026-02-26 15:58:27","publishedOnDateReadable":"February 26th, 2026"},"versionCreatedAt":"2025-09-01 09:26:25","video":"","vorDoi":"10.1038/s41598-026-41419-0","vorDoiUrl":"https://doi.org/10.1038/s41598-026-41419-0","workflowStages":[]},"version":"v1","identity":"rs-7191610","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7191610","identity":"rs-7191610","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}