Short-term impacts of TV viewing on executive functions in children with reading difficulties: an eye tracking study

Abstract

Background

Studies have shown that watching fantastic TV programs immediately impacts typically developing children’s executive functions (EFs). TV program contents may have a different effect on children with reading difficulties (RDs) because of executive dysfunctions, which have not been studied yet. This study examined the short-term effects of fantastical and realistic TV content on visual attention and inhibitory control in children with RDs and typically developing (TD) peers, considering roles of age and behavioral/emotional problems.

Methods

Forty-seven boys aged 7.5–9.5 years (23 RD, 24 TD) completed eye-tracking’s anti-saccade (inhibitory control) and gap tasks (visual attention) before and immediately after viewing a fantastical or realistic TV program. Children’s behavior/emotional problems were assessed using the Strengths and Difficulties Questionnaire (SDQ) as a covariate.

Results

Visual attention outcomes showed no significant differences across groups or conditions. However, significant interactions emerged between age, hyperactivity, and TV exposure, indicating that these factors influenced attention. For inhibitory control, significant decreases in accuracy were observed in the fantastical RDs and realistic TD groups, highlighting the nuanced effects of content type. The effects of TV viewing were further moderated by children’s age and behavioral/emotional problems.

Conclusions

Fantastical TV content appears to impact EFs, particularly inhibitory control, in children with RD. This suggests that processing demanding content may differentially tax already vulnerable EFs. Age and behavioral/emotional problems also significantly influence these effects, emphasizing the importance of considering individual differences in screen time research. These findings underscore the need for further research to explore the cognitive impacts of TV programs on neurodivergent children and to inform evidence-based media recommendations.

Background

Reading Difficulty (RD) is a specific learning disability that is characterized by persistent challenges in accurate or fluent word recognition, decoding, and spelling, leading to challenges in acquiring and applying academic skills [1, 2]. RD is among the most common learning disabilities, with an estimated prevalence of 5–20%, though rates vary across studies and countries [3–5]. For instance, a prevalence rate of 5.7% has been reported in the Iranian population [6].

Despite its primary impact on reading and phonological processing, the effects of RD extend beyond academics, affecting individuals’ behavioral, emotional, and social domains [7]. Children with RD have a higher likelihood of exhibiting behavioral challenges compared to typically developing peers, ranging from internalizing problems, such as anxiety and depression, to externalizing problems, such as hyperactivity and conduct problems [7–10]. Furthermore, the cognitive and linguistic challenges related to RD can also impact social interactions with peers, leading to social isolation and emotional distress [10, 11].

RD is also associated with broader cognitive impairments, particularly in executive function (EF) such as working memory [12, 13], attention [14, 15], and inhibitory control [16–19]. These deficits contribute to difficulties in meeting academic and daily life demands. Various theoretical frameworks have been proposed to account for the cognitive deficits linked to RD, including phonological awareness [20], auditory discrimination [21], visual attention [22], working memory [23], and automatization processes [18].

RD, EFs, and environmental influences

Children with RD often demonstrate reduced EF performance [17, 19, 24, 25]. EFs encompass a set of high-level cognitive processes essential for goal-directed behavior, including working memory, processing speed, attention, inhibitory control, and cognitive flexibility [26, 27]. These processes are critical for adaptive behavior and self-regulation, enabling individuals to plan, organize, and navigate their environment effectively [28, 29]. EFs deficits may make children with RD particularly vulnerable to environmental influences, such as screen exposure, that are already known to affect EFs in TD children (see Sect. 1.2 [30, 31]),. This vulnerability is especially relevant given the cognitive demands reading places on EF components.

The relevance of EFs deficits in this population is underscored by the significant demands reading places on attention, inhibition, and working memory [32, 33]. Research consistently shows that children with RD exhibit deficits in visual attention [14, 15, 34] and inhibitory control [16, 35–37] as EF components. Visual attention, a complex cognitive process, allows individuals to selectively focus on specific stimuli within their visual environment, facilitating effective decision-making and behavior aligned with personal goals [38, 39].

Impairments in visual attention are strongly associated with RD, with individuals often exhibiting challenges in areas such as visual attention span, visuospatial attention, and visual search [15, 16, 40–42]. These difficulties manifest as slower performance on tasks requiring integrating multiple visual features, such as visual conjunction searches, and a reduced capacity to process multiple stimuli simultaneously [43]. Moreover, deficits in visual temporal processing, particularly the ability to segregate rapidly presented stimuli, are frequently observed in individuals with RD [44]. These results suggest that visual attention deficits, independent of phonological processing issues, play a significant role in the development and manifestation of RD by impairing the ability to efficiently process and integrate visual information critical for reading [34, 45].

In addition, RD is associated with deficits in inhibitory control, which is a core EF. Inhibitory control enables individuals to suppress responses or thoughts that may interfere with goal-directed behavior [26, 46]. Inhibitory control is essential in suppressing irrelevant lexical information, such as the substitution of words with orthographic neighbors [35–37, 47–49] and word recognition and error management [50]. These inhibitory challenges are particularly evident in tasks requiring attentional disengagement, such as anti-saccade tasks, where RD individuals struggle to suppress reflexive eye movements to irrelevant stimuli [37, 49].

Given these existing deficits in visual attention and inhibitory control, children with RD may be particularly susceptible to the cognitive demands imposed by different types of television (TV)content. This potential for heightened sensitivity forms a key rationale for investigating how different types of TV content might uniquely affect the EFs of children with RD. This rationale is grounded in prior findings with TD children, where specific types of screen content have been shown to impair EF performance temporarily [30, 31, 51].

Television programs’ impacts on EFs

Digital devices have become ubiquitous in children’s lives, with 95.9% of preschoolers using them daily [52, 53]. Despite the growing prevalence of smartphones and tablets, TV remains a dominant source of screen exposure, accounting for a substantial portion of children’s screen time [53, 54]. The negative impact of excessive TV viewing on children has spurred extensive research into its various dimensions, particularly the content of TV programs. Among these, the influence of fantasy content has been a topic of significant discussion and investigation [51].

Research suggests that watching fantastical TV content, characterized by narratives that defy the laws of physics, can significantly impact children’s EFs [31]. Infants possess an innate understanding of how the world operates, and when these fundamental expectations are violated by fantastical elements, it demands greater cognitive effort [31, 55]. This increased processing is necessary as children attempt to reconcile the impossible events presented in fantasy with their existing knowledge of reality. This cognitive burden arises from the novelty of fantastical events, which require additional cognitive resources to be assimilated into existing mental frameworks. Furthermore, the frequent violations of real-world expectations within fantasy content can overload children’s cognitive processing capacity, potentially disrupting the learning of intended concepts [56, 57].

While the impact of various media on children’s development has been extensively studied, the role of fantasy content within TV programs remains a subject of ongoing debate. Lillard et al. [31] pioneered research on the impact of fantasy on children’s cognitive development, finding that exposure to fantasy content in early childhood can negatively affect EFs in 4- to 6-year-olds. Subsequent research has delved deeper into the specific EFs impacted by fantasy. A prominent area of inquiry has been the impact of fantasy content on inhibitory control, with multiple studies reporting negative outcomes. For instance, Jiang et al. [30] and Li et al. [58] observed significant reductions in inhibitory control among children exposed to fantasy programs. Similarly, Rhodes et al. [59] found that children who watched high-fantasy episodes, such as Little Einstein, performed worse on inhibitory control tasks than those exposed to low-fantasy content like Little Bill. These findings align with broader research by Fan et al. [60], who demonstrated that high-fantasy cartoons had a greater impact on younger children, with 4-year-olds most affected, followed by 5- and 6-year-olds. In another detailed examination, Li et al. [61] used tools such as eye-tracking and functional near-infrared spectroscopy (fNIRS) to provide physiological evidence for these cognitive disruptions, noting increased oxyhemoglobin levels in the prefrontal cortex of children exposed to fantastical content. However, Keşşafoğlu et al. [62] introduced a temporal nuance, showing that a 10-minute delay between viewing and testing could mitigate the negative effects on inhibitory control.

Despite these consistent findings, some studies present a more nuanced or even contradictory picture. Kostyrka-Allchorne et al. [63] observed a positive influence of fantasy on cognitive abilities, reporting that children aged 3.5 to 5 years scored higher in inhibitory control after watching brief fantastical videos compared to realistic ones. Additionally, Hinten [64] found no significant differences in inhibitory control among children aged 5 to 7 years exposed to fantasy versus realistic cartoons. Interestingly, this study revealed that children who read fantasy books performed better on inhibitory control tasks than those who watched fantasy cartoons, suggesting that the medium of fantastical content might play a crucial role. Conversely, children who read realistic books performed worse than their counterparts who watched realistic cartoons.

In addition to inhibitory control, research has delved into other EF components, such as attention. Kostyrka-Allchorne et al. [63] hypothesized that fantastical content could enhance orienting responses and bottom-up processing, potentially benefiting attentional tasks. However, their results showed no significant differences between fantasy and realistic groups, underscoring the need for further investigation. The mixed findings across studies highlight the complexity of the relationship between fantastical content and EFs development. While the predominant evidence points to a negative impact, particularly on inhibitory control, contextual factors such as content type, exposure duration, developmental stage, and individual differences appear to mediate these effects [51].

Furthermore, while most studies exploring the effects of fantastical content on EFs have been conducted in Western or Chinese contexts, emerging research indicates that cultural factors may influence EF development and media processing [61]. For instance, the results of a recent study showed that Iranian preschoolers performed better in inhibitory control tasks than German preschoolers, but underperformed in working memory and cognitive flexibility tests [65]. These findings suggest that media-related cognitive effects may not generalize uniformly across cultural contexts. Therefore, investigating the effects of fantastical content within the Iranian context can deepen our understanding of how cultural factors modulate media impacts on EF development.

The current study

As the existing literature indicates, the type of TV content children consume can influence their EFs, particularly in TD populations. These findings raise critical questions about how such TV content might affect children with RD, whose baseline EFs profile (particularly in visual attention and inhibitory control) are already compromised. To date, no studies have investigated the impacts of TV programs on children with RD’s EFs. Addressing this gap, the present study aimed to examine the short-term effects of fantastical and realistic TV programs on the visual attention and inhibitory control of children with RD in the Iranian population. This investigation is especially important given that children with RD may experience exacerbated effects of screen exposure, especially when exposed to cognitively demanding content like fantasy. Based on prior research, we hypothesize that children with RD will show greater declines in in inhibitory control and visual attention after watching fantastical content compared to TD peers.

Moreover, RDs are often not identified until a child shows clear signs of falling behind, usually when a noticeable gap in reading skills and experience emerges compared to their peers, typically around second grade or later [66, 67]. For this reason, our study focused on children in grades 2 to 4, an age range that falls within middle childhood. Given the developmental trajectory of EFs, which mature throughout childhood and adolescence [29, 68–70], age is a critical factor in media-related cognitive outcomes. Previous research, such as Fan et al. [60], has demonstrated that the effects of fantastical content on inhibitory control are age-dependent, with younger children showing greater susceptibility. Building on these findings, the second aim of this study was to examine the role of age in moderating the short-term effects of fantastical and realistic TV content on EF performance in children with RD. It is hypothesized that age moderates the effect of TV on EF.

Finally, in addition to EFs deficit, children with RD usually experience behavioral and emotional difficulties [7, 9, 10, 71, 72]. While these challenges may not always reach clinical thresholds, they can significantly impact cognitive development and academic performance [73]. Such difficulties may also influence how children with RD respond to fantastical content, potentially amplifying its effects on EF components. Therefore, the third aim of this study was to investigate whether pre-existing behavioral/emotional problems influence the impact of fantastical content on EF components in children with RD. We hypothesize that behavioral/emotional problems moderate the effect of TV programs on EF.

Method

Participants

This study included 53 elementary school boys in grades 2 to 4, enrolled in public schools in Tehran, Iran. Of these, 27 children were TD, and 26 were diagnosed with RD based on school and teacher reports, according to the Diagnostic and Statistical Manual of Mental Disorders: DSM-5-TR [1]. The information sheets, consent forms, and questionnaires were distributed to parents in schools. Informed consent was obtained from parents, and children provided assent before participating in the study. The inclusion criteria required participants to be boys with normal or corrected-to-normal vision, no history of psychological, neurological, or sensory-motor disorders, and enrollment in grades 2 to 4. RD is more commonly diagnosed in boys than in girls, with studies reporting a higher prevalence among male students. The reported boy-to-girl ratio ranges from approximately 1.5:1 to 3:1 in general populations [74, 75]. Therefore, the inclusion of boys aligns with the demographic profile of the population most affected by RD. For the RD group, a teacher-confirmed diagnosis of reading disorder was mandatory, and participants with co-occurring comorbidities were excluded. It was assumed that all children enrolled in general schools had an normal IQ. Exclusion criteria included an inability to complete the eye-tracking tasks, lack of cooperation, and outlier data (3-scaled standard deviation away from the mean), excluding six participants. The final sample comprised 24 TD children and 23 children with RD (see Table 1 for details), with the groups matched for age to ensure comparability. Ethical approval for this study was obtained from the Shahid Beheshti University Ethics Committee (IR.SBU.REC.1402.101).

Table 1.

Participant profiles: demographic characteristics and group differences

Measures		Groups				Group differences (ANOVA)
		Reading Difficulties		Typically Developing
		Fantastic	Realistic	Fantastic	Realistic
N		12	11	11	13	-
Age (months)	Mean standard deviation	103.54 6.26	104.72 6.48	107.00 4.00	105.38 3.33	F = 0.869, p = 0.465, ηp2 = 0.058
	Minimum - Maximum	89–111	92–111	101–113	101–111

Material

TV programs

We used the same TV programs (interventions) as those in Lillard et al. [31], which was the first to investigate the impact of fantastical content on children’s EFs. These specific episodes have also been used in subsequent research [59], allowing for consistency across studies. This choice helps maintain a controlled program’s features while enabling us to explore the effects of changes in other methodological aspects. For the fantasy condition, the episode “Flight of the Instrument Fairies” from the Little Einsteins series was chosen, while the episode “All Tied Up” from the Little Bill animated series was selected for the realistic condition. The Persian versions of these programs were professionally dubbed and edited to match the segments used in the original study.

A coding form was developed to identify and record fantasy features, and two of the authors independently coded these features based on their exact timing (minute and second) following a specific set of instructions. These instructions included a definition of a fantasy feature (see Sect. 1.2) and specified that repeated occurrences of the same event were counted only once, at their first appearance. To assess pacing, we followed the method used by Lillard et al. [31], who utilized a computer program that estimates scene changes based on pixel variation from frame to frame, with an 85% pixel change serving as the threshold for identifying a scene change. For our analyses, we used Shutter Encoder 18.5 (2024) to facilitate this process (Table 2). It is worth noting that other potentially influential factors, such as language complexity, emotional tone, or the number of characters, were not addressed in either Lillard et al. [31] or the present study. The fantasy episode, “Flight of the Instrument Fairies,” includes numerous fantastical elements, such as impossible transformations and magical scenarios. In contrast, the realistic episode, “All Tied Up,” depicts the everyday life of a 5-year-old boy. While primarily grounded in reality, this episode contains a single fantastical element: a pair of self-walking, talking shoes.

Table 2.

Characteristics of TV programs

TV program	Length	Fantasy rate	Number of fantasy events	Pacing rate
Little Einsteins	8:34	1.28	16	9
Little Bill	8:23	0.12	1	16

EFs tasks (Visual attention and inhibitory Control)

This study assessed visual attention and inhibitory control using pro-saccade and anti-saccade tasks adapted from Amestoy et al. [76]. The gap task (pro-saccade) was employed to evaluate attention switching and disengagement. Participants were required to quickly shift their gaze from a central fixation point to a peripheral target. The anti-saccade task, heavily reliant on inhibitory control, demanded participants to suppress the reflexive saccade towards the peripheral stimulus and instead look in the opposite direction (Fig. 1).

Fig. 1.

Gap and anti-saccade tasks design. The green moon serves as the central fixation point, while the red star represents the peripheral target. The white arrow indicates the participant’s gaze

Tasks procedure

The experiment utilized a 23-inch screen (1920 × 1080 resolution, 144 Hz refresh rate), a Lenovo laptop, and a GP3 HD Eye Tracker (Gazepoint, Canada) with a 150 Hz sampling rate. Participants were seated 65 cm from the eye tracker and underwent a nine-point calibration before each task. Each trial began with a green fixation point (0.5 × 0.5 cm) displayed for 2000–3500 ms, followed by a red star (0.5 × 0.5 cm) appearing 1000 ms later, positioned 24 cm to either side of the central point. The fixation point disappeared 200 ms before the peripheral stimulus appeared. In the anti-saccade task, participants were instructed to look away from the peripheral stimulus, which required greater cognitive effort and resulted in longer latencies than the pro-saccade task. The minimally delayed oculomotor response task, designed to isolate inhibition processes, shows increased saccade latency compared to prosaccades [77]. The difference between Anti-saccade and Gap latencies is referred to as the “Anti-saccade effect.”

This study extracted four oculomotor variables: latency, gain, anticipatory saccades, and erroneous saccades. Latency measured the time elapsed before the first non-anticipatory fixation, encompassing both reaction time and saccade duration. The gain was calculated as the ratio of the actual saccade distance to the theoretically required distance (24 cm). Anticipatory saccades were defined as early saccades occurring within 100 ms. Erroneous saccades were categorized as non-anticipatory saccades directed towards an incorrect location.

Parent questionnaire

Parents completed the Strengths and Difficulties Questionnaire (SDQ) to assess their children’s behavioral and emotional problems. The SDQ, developed by Goodman [78, 79], is a widely used psychological screening tool for children and adolescents aged 2–17. This brief questionnaire comprises 25 items distributed across five subscales: Emotional Problems, Conduct Problems, Hyperactivity/Inattention, Peer Problems, and Prosocial Behavior. The first four subscales are combined to produce a Total Difficulties Score, which reflects overall behavioral and emotional challenges, while the Prosocial Behavior subscale measures strengths. In this study, only the results related to behavioral and emotional problems were analyzed; the Prosocial Behavior subscale was excluded. Responses on the SDQ are recorded using a three-point Likert scale: “Not True,” “Somewhat True,” and “Certainly True,” which are used to calculate scores for each subscale. The SDQ has demonstrated satisfactory internal consistency (Cronbach’s alpha = 0.73 [79]) and good validity compared to the Child Behavior Checklist [80]. The Persian adaptation of the SDQ has also shown high levels of reliability, with reported Cronbach’s alphas ranging from 0.71 [81] to 0.85 [82].

Procedures

Prior to the study, teachers identified children with RD through consultation. Subsequently, informed consent was obtained from all participating children’s parents. Children then underwent a standardized 30-minute assessment protocol in a quiet classroom at their school. After a brief introduction, children engaged in a series of tasks. The session began with visual attention and inhibitory control tasks, followed by an eight-minute viewing of one of two pre-selected TV programs. Subsequently, they repeated the initial attention and inhibition tasks. Upon completion, each child received a small reward, such as an eraser or pencil (Fig. 2). The study utilized standardized equipment, including a 23-inch screen, a Lenovo laptop, and a GP3 HD Eye Tracker to collect gaze data. To ensure stable measurements, children used a chinrest during testing. The tasks included a gap task to assess visual attention and an anti-saccade task to evaluate inhibitory control. A nine-point calibration procedure was conducted before each task, followed by four practice trials to ensure familiarity with the procedure. Each task included four practice trials that were identical in format to the experimental trials (Fig. 1). After completing the practice trials, children were asked whether they fully understood the task and if they had any questions. Following this, participants completed 15 experimental trials for each task. The data from the practice trials were excluded from the final analyses. The order of tasks was randomized for each participant to minimize order effects, with each task lasting approximately four minutes. These descriptions were largely drawn from our previous study, which employed comparable methods.

Fig. 2.

Study design: group assignment and intervention procedure

Data processing and analytic strategy

Eye movement data were collected using an eye-tracking device, following the methodology outlined by Amestoy et al. [76]. Consistent with their approach, we examined eye fixation parameters, including X and Y coordinates, latency, accuracy, stimulus presentation time, and fixation duration on both central and peripheral stimuli. Data analysis was conducted using R programming (version 4.3.1). Initially, trials with low-quality or invalid data were excluded, as were trials where participants failed to fixate on the central point within a 4 cm tolerance immediately preceding the target appearance. Latency and gain were subsequently calculated for non-anticipatory, non-erroneous saccades. Trials with first saccade gain values less than one-third of the target distance or exceeding three standard deviations were excluded from further analysis. The effects of different TV program groups on visual attention and inhibitory control in both RD and TD children were analyzed using mixed-design ANOVAs. After that, we applied the Bonferroni adjustment method exclusively for pairwise comparisons to reduce the risk of inflated Type I error. Mixed-design ANCOVAs were performed to further examine the influence of age and behavioral/emotional problems. When a significant effect emerged in the initial analysis, follow-up grouped ANCOVAs were performed to explore the main effects in more detail. Prior to conducting ANCOVA analyses, assumptions were rigorously tested, and alternative analytical methods were employed when these assumptions were violated. For variables that violated the normality assumption (based on visual inspection and formal tests), we adopted Generalized Linear Mixed Models (GLMMs), which are appropriate for non-normal outcome distributions from the exponential family (e.g., Poisson, binomial). GLMMs also allowed us to account for subject-level random effects and were particularly suitable for modeling repeated-measures data. In addition, we employed a grouped ANCOVA approach for some variables that were normally distributed but violated the assumption of homogeneity of regression slopes (i.e., a significant covariate-by-group interaction). In these cases, the data were divided into meaningful subgroups based on the variable and interaction where the violation occurred, and ANCOVA was performed separately within each group to satisfy the assumption. GLMMs were fitted using a gamma (skewed) distribution with a log link function for non-normally distributed data and a Gaussian distribution with an identity link function for normally distributed data. Model selection was guided by the Akaike Information Criterion (AIC). In the GLMMs, session, program group, and children’s group were included as fixed effects, age and behavioral/emotional problems were entered as covariates, and participant ID was specified as a random intercept to account for within-subject variability. We used Type III Wald chi-square tests for evaluating fixed effects.

Results

Preliminary analyses

Descriptive statistics for behavioral and emotional problems within each TV program and children’s group are presented in Table 3. A series of one-way ANOVAs revealed statistically significant differences in hyperactivity, peer problems, and total behavioral/emotional problems (total SDQ). Specifically, children in the RD groups exhibited significantly higher levels of hyperactivity compared to other groups. Additionally, children in the Fantastic RD group showed significantly higher levels of peer problems and total behavioral/emotional problems compared to children in the TD groups.

Table 3.

Descriptive statistics and group comparisons of behavioral/emotional problems at study onset

Variables		Groups (meanstandard deviation)				Group differences (ANOVA)
		RD		TD
		Fantastic	Realistic	Fantastic	Realistic
SDQ	Emotional problems	3.001.56	2.402.17	1.5451.036	2.231.16	F = 1.650, p = 0.194, ηp 2 = 0.11
	Conduct problems	2.801.57	2.002.26	1.3641.43	2.302.17	F = 1.02, p = 0.395, ηp 2 = 0.071
	Hyperactivity	5.502.79	5.102.64	2.541.81	2.421.93	F = 5.40, p = 0.003**, ηp 2 = 0.294
	Peer problems	4.201.32	2.701.76	2.001.73	1.611.56	F = 5.470, p = 0.003**, ηp 2 = 0.291
	Total	15.506.06	12.205.61	7.452.87	8.353.88	F = 6.660, p = 0.001**, ηp 2 = 0.333

Visual attention results

A series of mixed-design ANOVAs were conducted to examine the effects of the TV program group, children's group, and session on the variables of latency, gain, anticipatory saccades, and erroneous saccades. The analyses revealed no significant effects for the session, programs, children's group, or their interactions on any of the visual attention variables. The main effect of the program group was not significant for latency (Fig 3A; F = 0.400, p = 0.531, η_p² = 0.009), gain (Fig 3B; F = 0.642, p = 0.427, η_p²= 0.015), anticipatory saccades (Fig 3C; F = 0.754, p = 0.390, η_p²= 0.017), and erroneous saccades (Fig 3D; F = 1.065, p = 0.308, η_p²= 0.024). Similarly, no significant main effects of the children's group or session were observed for latency (Fig 3A; F = 3.092, p = 0.086, η_p²= 0.067; F = 0.355, p = 0.554, η_p² = 0.008), gain (Fig 3B; F = 0.008, p = 0.929, η_p²= 0.0002; F = 0.632, p = 0.431, η_p² = 0.014), anticipatory saccades (Fig 3C; F = 0.542, p = 0.466, η_p²= 0.012; F = 0.182, p = 0.672, η_p² = 0.004), and erroneous saccades (Fig 3D; F = 0.007, p = 0.932, η_p²= 0.0002; F = 0.514, p = 0.477, η_p² = 0.012). Furthermore, none of the interactions between the TV program group, children's group, and session were significant for any of the visual attention variables (all p > 0.05).

Fig. 3.

Impact of TV Program Features on Visual Attention Variables in Different Child Groups. The figure illustrates the impact of TV program type (Realistic vs. Fantastic) on four components of visual attention: A) Latency, B) Gain, C) Anticipatory Saccades, and D) Erroneous Saccades across pre-test and post-test sessions. No significant differences were observed between sessions, program types, or groups (RD and TD), even after controlling for covariates such as age and behavioral/emotional problems. However, there was a significant effect of session and hyperactivity on gain within the fantasy group, which was not observed when the fantasy group was divided into TD and RD subgroups

Age

ANCOVA results revealed a significant effect of age on anticipatory saccade (Fig 4; F = 4.601, p = 0.038, η_p² = 0.101), however, there were no significant age-related impacts on latency (F = 0.248, p = 0.621, η_p² = 0.006) and gain (F = 0.559, p = 0.459, η_p²= 0.013). Similarly, no significant effects of the TV program, children's group, or session were observed on gain or anticipatory saccades, even after controlling for age (all p > 0.05). However, a significant effect of the children’s group and session on latency was observed after controlling for age (F = 4.084, p = 0.050, η_p² = 0.091; F = 4.895, p = 0.033, η_p² = 0.107). Additionally, a significant interaction between age and session was identified (F = 4.754, p = 0.035, η_p²= 0.104). For further analysis of the significant main effect, a grouped ANCOVA was conducted. The analyses revealed no significant effect of age on latency in either the pre-test (F = 0.335, p = 0.566, η_p²= 0.008) or the post-test (F = 2.180, p = 0.147, η_p²= 0.051). However, the children’s group effect was significant in the post-test (F = 5.590, p = 0.023, η_p² = 0.120). Erroneous saccade did not meet the normality and homogeneity of regression slope assumptions required for ANCOVA. Therefore, a generalized linear mixed model (GLMM) was employed to address the issues of non-normally distributed residuals and the violation of the homogeneity of regression slope assumption. The analysis revealed no significant effects of the TV program group (χ² = 0.389, p = 0.532), children’s group (χ² = 0.041, p = 0.838), or session (χ² = 0.213, p = 0.644) on erroneous saccades after controlling for age. Furthermore, no significant effect of age was observed (χ²= 0.118, p = 0.731). Effect sizes were estimated as conditional R² values [83], representing the proportion of variance explained by the full model. The conditional R² was 0.044, indicating that approximately 4.4% of the variance in the outcome was explained by the model.

Fig. 4.

Effect of age on visual attention anticipatory saccades across program types and groups. Scatter plots showing the relationship between age and anticipatory saccades

Behavioral/emotional problems

Several ANCOVAs were conducted to examine the effect of different components of behavioral and emotional problems on visual attention. The results indicated that none of the subscales of behavioral/emotional problems had significant impact on latency (emotional problems: F = 0.383, p = 0.540, η_p²= 0.010; hyperactivity: F = 0.872, p = 0.356, η_p²= 0.022; peer problems: F = 0.344, p = 0.561, η_p²= 0.009; total SDQ: F = 0.145, p = 0.705, η_p²= 0.004). Additionally, after controlling for each of these subscales, no significant effects of the TV program group, children’s group, or session were observed for latency (all p > 0.05). However, due to the violation of the homogeneity of regression slope assumption for conduct problems, grouped ANCOVAs were conducted, revealing a significant effect of conduct problems on latency within the Realistic RD group (Fig 5A; F = 8.930, p = 0.017, η_p²= 0.528). For the gain variable, the results were similar, with no significant effects found for any subscale (emotional problems: F = 0.111, p = 0.741, η_p² = 0.003; conduct problems: F = 0.263, p = 0.611, η_p²= 0.007; total SDQ: F = 1.328, p = 0.256, η_p²= 0.033). However, due to the violation of the homogeneity of regression slope assumption for the hyperactivity and peer problems subscales, grouped ANCOVAs were used. An effect of hyperactivity (F = 4.690, p = 0.044, η_p²= 0.207), session (F = 7.930, p = 0.011, η_p²= 0.306), and the hyperactivity and session interaction (F = 6.560, p = 0.020, η_p²= 0.267) on the children's gain score was observed in the fantasy group but not in the realistic group (all p > 0.05). This indicates that the impact of repeated exposure (pre- to post-test) on visual attention gain was modulated by children's hyperactivity levels. However, when the fantasy group was separated into TD and RD subgroups, this interaction effect was no longer significant. Additionally, no significant effects were observed for peer problems (all p > 0.05).

Fig. 5.

Effect of behavioral/emotional problems on visual attention components across program types and groups. Scatter plots illustrate the relationship between total behavioral/emotional problems, conduct problems, hyperactivity, and three visual attention components: A) Latency. B) and C) Anticipatory Saccades, and D) and E) Erroneous Saccades

For anticipatory saccades, a significant effect of the total behavioral/emotional problems (SDQ) score (Fig 5B; F = 7.178, p = 0.011, η_p²= 0.155) and conduct problems (Fig 5C; F = 4.873, p = 0.033, η_p²= 0.111) on the anticipatory saccades were observed. However, after controlling for individual subscales, there were no significant effects of the program group, children’s group, or session (all p > 0.05). Finally, due to non-normality in the erroneous saccade variable, a GLMM was used. Additionally, GLMM was employed for the hyperactivity subscale due to the violation of the homogeneity of regression slope assumption. The results demonstrated significant effects of the total behavioral/emotional problems score and the hyperactivity subscale on erroneous saccades (Fig 5D, E; hyperactivity: χ² = 6.428, p = 0.011; total SDQ: χ² = 6.245, p = 0.012). However, after controlling for each subscale, no significant effects of the program group, children’s group, or session were observed (all p >0.05). The conditional R² was 0.207 and 0.185 for models with hyperactivity and total behavioral/emotional problems, respectively.

Inhibitory control results

Mixed-design ANOVAs were conducted to examine the effects of the TV program group, children's group, and session on inhibitory control variables, including latency, gain, anticipatory saccades, erroneous saccades, and the anti-saccade effect. The analyses revealed no significant effects of the children's group on any of the variables (Fig 6; F = 0.994, p = 0.324, η_p²= 0.023; F = 0.045, p = 0.832, η_p² = 0.001; F = 0.336, p = 0.565, η_p² = 0.008; F = 0.061, p = 0.806, η_p² = 0.001; F = 0.081, p = 0.777, η_p² = 0.002). Similarly, the program group showed no significant main effects on latency (F = 0.085, p = 0.772, η_p² = 0.002), gain (F = 0.524, p = 0.473, η_p²= 0.012), erroneous saccades (F = 1.990, p = 0.166, η_p²= 0.044), and anti-saccade effect (F = 0.307, p = 0.582, η_p²= 0.007). However, a significant main effect of the program group was observed for anticipatory saccades (F = 4.896, p = 0.032, η_p²= 0.102). Additionally, unlike the other variables (all p > 0.05), gain demonstrated a significant effect of the session (F = 7.633, p = 0.008, η_p²= 0.151) and interaction of session, TV program group and children's group (F = 8.233, p = 0.006, η_p² = 0.161), suggesting differential dynamics across sessions. The post hoc analysis for the program group effect on anticipatory saccades revealed a significant effect in the pre-test (F = 4.260, p = 0.045, η_p²= 0.090), but no significant differences were observed in the post-test or across different groups (all p > 0.05). For gain, a main effect analysis identified a significant interaction effect of TV program group and children's group in the post-test (F = 6.900, p = 0.012, ηp2 = 0.138). The post hoc pairwise tests revealed that children with RD which exposure to fantastic programs showed significantly lower gain scores than the realistic group (t = -2.810, p = 0.011, d = -1.169). Also, there was a significant difference between the pre- and post-test scores within the Fantastic RD (Fig6B; t = -3.710, p = 0.003, dz = -1.07) and Realistic TD groups (Fig 6B; t =-3.530, p = 0.004, dz = -0.978).

Fig. 6.

Impact of TV program features on inhibitory control variables in different child groups. The figure illustrates the impact of TV program type (Realistic vs. Fantastic) on five components of inhibitory control: A) Latency, B) Gain, C) Anticipatory Saccades, D) Erroneous Saccades, and E) Anti-saccade Effect across pre-test and post-test sessions. Panel (A) demonstrates a significant interaction effect of session and conduct problems on latency within the Realistic TD group (indicated by *). However, this effect was not significant without controlling for conduct problems. Panel (B) shows a significant effect of the session on gain within the Realistic TD and Fantastic DYX groups, even after controlling for all covariates (indicated by *). Panel (C) highlights a significant interaction effect of session and emotional problems on anticipatory saccades within the Realistic TD group after controlling for emotional problems (indicated by *). However, this effect was not significant without controlling for emotional problems. In panel (D), the erroneous saccade in the Fantastic DYX group shows a significant interaction effect of the session and both age and hyperactivity (indicated by *). This effect was not significant without controlling for these covariates. Panel (E) reveals no significant differences between groups for the anti-saccade effect

Age

Since all residuals were normally distributed, ANCOVA was used to examine the role of age across all variables. However, due to violations of the homogeneity of regression slope assumption, grouped ANCOVA was applied for latency, erroneous saccades, and the anti-saccade effect. The results revealed that the significant interaction effect on gain (F = 6.711, p = 0.013, η_p²= 0.141) and the program group's effect on anticipatory saccades (F = 4.210, p = 0.047, η_p² = 0.093) persisted after controlling for age. These findings aligned with previous results, demonstrating significance for gain in both the Fantastic RD (Fig 6B; F = 9.730, p = 0.012, η_p² = 0.520) and Realistic TD groups (Fig 6B; F = 14.800, p = 0.003, η_p²= 0.574). However, the significant effect of the program group on anticipatory saccades observed in the pre-test disappeared (F = 3.800, p = 0.058, η_p²= 0.085). Moreover, a significant effect of age was observed exclusively in the Fantastic TD group for latency (Fig 7A; F = 18.000, p = 0.002, η_p²= 0.667) and the anti-saccade effect (Fig 7B; F = 10.800, p = 0.009, η_p²= 0.547). For erroneous saccades, the interaction of age and session was significant in the Fantastic RD group (Fig 6D; F = 5.190, p = 0.049, η_p²= 0.366), underscoring the role of age in the influence of the fantasy program on the RD group. However, no significant interaction was found in the Realistic RD (F = 4.540, p = 0.062, η_p² = 0.335), Realistic TD (F = 0.233, p = 0.639, η_p²= 0.021), or Fantastic TD (F = 0.642, p = 0.444, η_p²= 0.067) groups.

Fig. 7.

Effect of age on inhibitory control components across program types and groups. Scatter plots showing the relationship between age and two visual attention components: A) Latency and B) Anti-saccade Effect

Behavioral/emotional problems

As in the previous section, normality was assumed for all variables. For latency and the anti-saccade effect, violations of the homogeneity of regression slope assumption were observed across all subscales of behavioral and emotional problems. Therefore, grouped ANCOVA was used, except for hyperactivity in latency. For the gain variable in the conduct problems, hyperactivity, and total behavioral/emotional problems components, a violation of the homogeneity of regression slope assumption led to the use of GLMM, as grouped ANCOVA could not be applied. The results for latency indicated that neither the program group nor the children's group had a significant effect after controlling for all subscales of behavioral/emotional problems (all p > 0.05). However, in the Realistic TD group, a significant interaction between session and conduct problems was found for latency (Fig 6A; F = 7.600, p = 0.019, ηp² = 0.409), indicating that changes in latency from pre- to post-test were moderated by the level of conduct issues. This effect was not observed in the other groups (all p > 0.05). A significant three-way interaction was observed among program group, children’s group, and session on the gain variable, especially in the post-test (all p < 0.05). The pre and post-test difference was significant in both the Fantastic RD (Fig 6B; emotional problems: F = 10.300, p = 0.012, η_p² = 0.562; hyperactivity: F = 10.700, p = 0.011, η_p²= 0.572) and Realistic TD (Fig 6B; emotional problems: F = 11.400, p = 0.006, η_p²= 0.510; hyperactivity: F = 8.960, p = 0.013, η_p²= 0.473) groups after controlling for all covariates.

For anticipatory saccades, similar to previous findings, a significant effect of the program group persisted even after controlling for emotional problems, conduct problems, and total behavioral/emotional problems. However, the initial pre-test significance did not hold after controlling for the individual subscales. Emotional problems (Fig 8A; F = 5.350, p = 0.049, η_p²= 0.401) significantly affected anticipatory saccades in the Fantastic RD group, with an interaction between this subscale and session observed in the Realistic TD group (Fig 6C; F = 5.680, p = 0.036, η_p²= 0.341). Regarding erroneous saccades, a significant session effect was found after controlling for emotional problems (F = 5.876, p = 0.020, η_p²= 0.131), hyperactivity (F = 7.077, p = 0.011, η_p²= 0.157), and total behavioral/emotional problems (F = 5.399, p = 0.025, η_p²= 0.122), with an interaction between session and hyperactivity (F = 5.538, p = 0.024, η_p² = 0.127). Further analysis showed that only the total behavioral/emotional problems score significantly impacted the post-test percentage of erroneous saccades (Fig 8B; F = 5.860, p = 0.020, η_p²= 0.131). A significant interaction between session and hyperactivity was found for erroneous saccades in the Fantastic RD group (Fig6D; F = 6.960, p = 0.030, ηp² = 0.465). No significant effects were observed in the other groups or subscales (Fig 6E; all p > 0.05). Finally, grouped ANCOVA results indicated no significant session effect on the anti-saccade effect in any group, even after controlling for behavioral/emotional problems (Fig 6E; all p > 0.05).

Fig. 8.

Effect of behavioral/emotional problems on inhibitory control components across program types and groups. Scatter plots illustrate the relationship between total behavioral/emotional problems, emotional problems, and two visual attention components: A) Anticipatory Saccades, and B) Erroneous Saccades

Discussion

Previous research has demonstrated that exposure to fantastical TV programs can negatively impact EFs in TD children [51]. This study examined the effects of both fantastical and realistic TV content on visual attention and inhibitory control in children with RD.

The impact of fantastic and realistic TV Programs on visual attention

A key finding of our study was that viewing 8-minute segments of fantastic or realistic TV programs did not yield a significant main effect on the visual attention of children with or without RD, as measured by the gap task. This result aligns with Kostyrka-Allchorne et al. [63], who similarly reported no significant impact of fantasy content on children's attention. However, it contrasts with the findings of Portugal et al. [84], which demonstrated the effects of digital device exposure on children's attention, as well as with studies by Lillard et al. [31] and Li et al. [61], which reported an influence of fantasy content on children's EFs.

The lack of observed differences in visual attention in this study may be attributed to the nature of the gap task used in this study, which primarily assesses bottom-up visual processing [84]. Fantastical content, which often includes strange or unexpected elements, may place more demands on higher-level processes, like cognitive control and working memory, as viewers try to make sense of new and unfamiliar information. This means children need to use more working memory to keep track of unfamiliar characters, unusual settings, and unexpected rules [85]. According to Lang's limited capacity model, fantastical content requires a greater engagement of top-down cognitive resources to integrate and interpret unusual elements [57], potentially reducing reliance on bottom-up attentional processes [51]. However, if the fantastical stimulus isn't sufficiently demanding or the EF task isn't sensitive to the specific cognitive processes being taxed, null effects may emerge.

The gap task is mainly used to measure basic visual attention, especially how quickly a person can shift their gaze from one point to another (saccadic reaction time) and how easily they can disengage their attention from a central stimulus [86, 87]. This could explain why simpler attention tasks like the gap task may not be sensitive enough to pick up on these effects. It’s also possible that fantastical content doesn’t affect basic visual attention at all but instead influences more complex abilities, like sustained attention (keeping focus over time) or attentional flexibility (shifting between different types of information). These higher-level functions may be more vulnerable to the mentally demanding nature of fantastical stories. Therefore, future studies could benefit from incorporating tasks that assess these more complex attentions to determine whether fantastical content has more specific effects on attention.

Moreover, watching just 8 minutes of a TV program may not be enough to cause noticeable short-term changes in a relatively stable cognitive skill like basic visual attention. It’s possible that longer or repeated exposures are needed to see clearer effects on attention, especially in children who are already accustomed to frequent media use. The age range of participants in this study (7.5–9.5 years) is another critical consideration. Developmental changes in visual attention and cognitive processing during early childhood may influence how children respond to TV content [60]. Previous studies have shown that attentional processes and EFs undergo significant growth during the preschool period [88].

On the other hand, due to not using a standardized test to identify children with RD, it’s possible that some participants were not accurately classified. As a result, even if some of these children experience attention-related challenges, the lack of a clear effect from any of the programs might reflect problems with how the RD group was defined rather than a true absence of impact. Additionally, the gap task may not fully capture the range of visual attention deficits typically associated with RD, which often involve difficulties in integrating multiple visual features and maintaining attention over extended periods [15, 16, 40, 41]. Future research should utilize a broader array of attentional assessments to comprehensively examine the relationship between visual attention deficits and RD in children. Incorporating standardized diagnostic tools to better characterize RD populations may also enhance the reliability and interpretability of such findings.Our investigation examined the role of individual characteristics, specifically age and behavioral/emotional problems, within the context of exposure to fantastical TV content. Key findings revealed a significant interaction between session (pre-test vs. post-test) and age on visual attention latency. This interaction suggests that the rate of change in visual attention may be varied across age ranges during the study period. This result is consistent with findings from Fan et al. [60], which demonstrated that age influences the outcomes of TV exposure in children. These findings can be related to the interplay between age and TV exposure in shaping attentional development, aligning with established developmental trajectories that indicate attentional speed improves with age during specific developmental windows [89].

The study also uncovered that the impact of fantasy exposure on visual attention accuracy depends on a child's level of hyperactivity. This aligns with hypotheses suggesting that pre-existing behavioral challenges like hyperactivity significantly affect EFs and can moderate responses to cognitive demands [73]. Notably, the significant effects disappeared when the fantastical group was further divided into TD and RD subgroups. This suggests that the initial interaction may not have been primarily driven by reading ability status, but rather by a more complex interplay between hyperactivity and exposure to fantastical content. While children with RD exhibited significantly higher hyperactivity scores compared to their TD peers, the absence of significant effects in subgroup analyses may reflect limited statistical power due to smaller sample sizes. Therefore, these findings should be interpreted with caution, and future studies with larger and more balanced samples are needed.

The impact of TV programs on inhibitory control

The effects of TV programs on inhibitory control varied across different components. No significant impact was observed on latency, erroneous saccade, or anti-saccade effect, aligning with Hinten [64] but contrasting with studies that reported inhibitory control impairments following fantastical TV exposure [30, 31, 60–62]. These discrepancies may be due to the age of participants in this (7.5-9.5 years) and Hinten (5-7 years [64]) studies, as the sensitive period for inhibitory control development is between 3-6 years [70, 90]. Older children may have stronger baseline inhibitory control skills, which could make them less vulnerable to the immediate and subtle effects of short-term exposure to fantastical content, especially when it comes to well-established inhibitory responses [60]. From a cognitive load perspective [91], although fantastical content tends to be demanding, the specific processes measured by the anti-saccade task may be relatively resistant to the general cognitive load caused by passive media viewing in this age group. This resilience likely holds unless the cognitive load is particularly high or directly interferes with the mechanisms involved in selecting and controlling responses.

Additionally, the use of a fixed task for both pre- and post-testing may have introduced a learning effect, potentially masking the impact of TV program content [92]. Given the role of novelty in engaging EFs, repeated exposure to the same task might have diminished the observed effects. The last factor to consider is culture. For instance, previous findings suggest that Iranian children demonstrate higher levels of inhibitory control compared to their German counterparts [65]. This cultural difference may help explain the current results, as children with stronger inhibitory control may be more resistant to short-term influences from media exposure. Future research should explore this possibility through cross-cultural comparisons to better understand how cultural background may shape children’s susceptibility to media effects.

However, accuracy (gain) scores decreased from pre- to post-test in both fantastic RD and realistic TD groups. This partially aligns with previous research showing decreased inhibitory control after exposure to fantastical programs [30, 31, 60–62]. The findings support the assumption that children with RD, who often exhibit underlying EF vulnerabilities [73], may be more susceptible to the cognitive demands or disruptive effects of certain media content. For children with RD, who may have more limited working memory capacity or rely on less efficient information processing strategies [93], the increased cognitive demands of fantastical content could quickly drain their resources. As a result, their performance on a challenging EF task like the anti-saccade may suffer, reflected in lower accuracy scores.

Moreover, the decrease in the realistic TD group contrasts with some studies that found positive effects of realistic programs on EFs of TD children [31, 58]. This inconsistency might be attributed to the anti-saccade task, which specifically measures inhibitory control and may be more sensitive to TV exposure effects than other tasks. Unlike other measures, the anti-saccade task requires suppressing a reflexive response and generating a voluntary response, which may be uniquely influenced by exposure to fantasy content. Another critical consideration is the interpretation of accuracy scores in the gain component. The observed decrease in gain scores could be related to slightly improved latency performance, suggesting a potential shift in cognitive resource allocation rather than global impairment. The observed decrease in gain scores could be indirectly related to slightly (non-significantly) improved latency performance, indicating minimal improvement. This suggests that while accuracy may decline, it may not necessarily reflect a global impairment but rather a shift in the focus of cognitive resources. Furthermore, depending on the realistic content nature (if it was overly familiar), it may have led to a state of under-arousal or mild boredom in TD children. Suboptimal arousal can impair performance on tasks that demand strong attentional control and precision [94], such as the anti-saccade task. Additionally, significant effects of TV programs on anticipatory saccades at pre-test likely stem from small baseline differences between groups.

After controlling for age, the observed effect on inhibitory control gain persisted across groups, while the pre-test effect of the TV program group on anticipatory inhibitory control disappeared. This suggests that the TV program's impact on inhibitory control gain may operate independently of age-related changes. In contrast, age served as a confounding variable in the anticipatory inhibitory control effect, which is consistent with previous research. Younger individuals typically exhibit better anticipatory inhibitory control due to the enhanced efficiency of their EF and attentional systems [95]. Additionally, an interaction effect between age and session on erroneous inhibitory control was observed in the fantastical RD group. This finding aligns with research indicating that age influences the outcomes of TV exposure in children [60]. The interaction may likely reflect the ongoing maturation of inhibitory processes, which vary among individuals with RD. This variation may also indicate differing rates of development in inhibitory processes related to reading skills across different ages. This underscores the need to account for age when interpreting the impact of TV content on children with RD.

The analysis also revealed that the effects on inhibitory control gain and anticipatory saccade inhibitory control remained consistent after controlling for behavioral and emotional problems subscales. This suggests an isolated impact of TV program exposure on inhibitory control accuracy, with behavioral and emotional problems acting as confounding factors, particularly for anticipatory saccades. Furthermore, specific interactions between session and conduct problems affected inhibitory control latency, while session and emotional problems influenced anticipatory saccade inhibitory control in the realistic TD group. These results suggest that conduct and emotional problems can modulate the effectiveness of inhibitory control in TD children exposed to realistic TV content. For TD children who exhibit more conduct or emotional problems, exposure to realistic TV content may have influenced their responses differently. Even though the content was not fantastical, it might have subtly altered their arousal levels or triggered emotional sensitivity, which in turn could have affected their performance on the subsequent task [94, 96].

After accounting for emotional problems, hyperactivity, and total behavioral/emotional problems, the main effect of the session on inhibitory control erroneous saccades was observed. This finding indicates that these factors may act as potential confounders in the relationship between TV program exposure and inhibitory control, irrespective of content type. The role of behavioral and emotional problems in moderating this effect is consistent with prior research, such as Plomin et al. (2002), which demonstrated associations between behavioral problems and deficits in EF. These findings underscore the importance of considering children’s behavioral and emotional problems when examining the influence of TV content on cognitive development.

Moreover, an interaction effect between session and hyperactivity on erroneous saccade inhibitory control was identified within the fantastical RD group. This highlights a potentially complex interplay between hyperactivity, fantasy content, and inhibitory control in children with RD. Specifically, it suggests that children with varying levels of hyperactivity respond differently to fantasy content, affecting their ability to inhibit erroneous responses. This may be explained by the role of hyperactivity, which can further impair the ability to sustain attention and increase errors on tasks requiring response inhibition [97]. Children with higher levels of hyperactivity also seek greater stimulation [98], making fantastical content especially engaging, but more distracting. This heightened arousal could interfere with their cognitive control and reduce their ability to inhibit prepotent responses. Additionally, children with RD in this study scored significantly higher in hyperactivity, a finding consistent with previous research [7, 10]. These studies demonstrate that children with RD frequently exhibit behavioral problems and EF deficits. Hyperactivity, in particular, plays a significant role in cognitive control, especially in the context of fantastical content and RD. This emphasizes the importance of considering comorbid conditions, such as hyperactivity, when studying cognitive control and its relationship with media exposure in children.

The influence of age and behavioral/emotional problems on EF

In addition to TV exposure, our results demonstrated significant effects of age and behavioral/emotional problems on visual attention and inhibitory control. With increasing age, we observed improvements in visual attention, specifically an increase in the percentage of anticipatory saccades. This aligns with previous studies [99, 100], which have shown that anticipatory (or predictive) saccades develop gradually during childhood, typically reaching adult-like levels around age 16. Anticipatory saccades reflect the ability to predict the timing and location of a target based on learned patterns, a skill that matures as EF and attentional control improve with age.

Similarly, we observed a decrease in inhibitory control latency and the anti-saccade effect with increasing age, particularly in the fantastical TD group. This suggests that older children may demonstrate faster response times and better control over reflexive eye movements, indicative of more advanced inhibitory control [101]. These improvements are attributed to the natural maturation of the prefrontal cortex, which governs executive functions such as inhibitory control, as supported by previous research [29, 68, 69]. The fact that these effects were significant only in one group may reflect individual differences and uncontrolled factors within this study, such as variability in cognitive development, exposure to other media, or baseline behavioral characteristics.

In relation to behavioral/emotional problems, our findings showed that as the total score of these problems increased, the percentage of anticipatory saccades decreased, while the percentage of erroneous saccades in visual attention increased. Similarly, a specific increase in behavioral problems was also associated with a decrease in the anticipatory saccades. This pattern suggests a decline in visual attention performance with higher levels of behavioral/emotional problems. Specifically, as behavioral problems increased, the percentage of anticipatory saccades declined, highlighting the detrimental impact of behavioral challenges on predictive attentional processes. Children with higher behavioral problems might consciously overcontrol their impulses to compensate for errors, reducing anticipatory saccades (preemptive eye movements). However, this excessive effort could exhaust cognitive resources, leading to more errors in visual attention tasks [102].

On the other hand, increased emotional problems were associated with an increase in anticipatory saccade inhibitory control, but this effect was significant only in the fantastical RD group. This finding suggests a decrease in inhibitory control performance with increasing emotional problems. The significant effects observed only in this group may be explained by slightly higher emotional problem scores in the fantastical RD group compared to other groups, although this difference was not statistically significant. These results are consistent with the findings of Plomin et al. [73], which demonstrated a relationship between EF deficits and behavioral/emotional problems.

Lastly, our results indicated that as the total score of behavioral/emotional problems increased, the percentage of inhibitory control erroneous saccades in the post-test also increased. This finding demonstrates a decline in inhibitory control performance after general TV program exposure among children with high scores in behavioral/emotional problems. These results support our hypothesis and highlight the vulnerability of EFs in children with elevated behavioral/emotional problems. This vulnerability may exacerbate difficulties in adapting to external stimuli, such as media content, further impacting cognitive performance.

Limitation

Like all research, this study has several inherent limitations. Initially, participant selection relied solely on teacher and school reports to identify children with RD. While such reports offer valuable ecological insights into children's learning challenges, they do not have the psychometric precision of standardized reading assessments or comprehensive neuropsychological evaluations. This method may have introduced variability in the diagnostic profile of the RD group. Future studies should incorporate clinical assessments and standardized psychometric tests to confirm RD diagnoses, enhance participant identification accuracy, and potentially explore the relationship between cognitive functions within RD and media effects. Additionally, the study exclusively included male participants, primarily due to the higher reported prevalence of learning disabilities in this group and considerations of accessibility in this study. While this controlled for potential gender-based variance, it significantly restricts the generalizability of the findings to female children, who may exhibit different EF profiles or responses to media. Future research should strive for more diverse samples, including both male and female participants.

Furthermore, the relatively small sample size in each group, particularly when conducting subgroup analyses (e.g., TD vs. RD within different exposure conditions), likely limited statistical power. This may have hindered the ability to detect more subtle effects or significant interactions. The combination of a modest sample size with multiple statistical comparisons and subgroup analyses also raises concerns about an increased risk of Type I errors and potential variance inflation. Additionally, family-wise adjustments such as the Bonferroni correction were not applied to the ANOVA and ANCOVA analyses. Therefore, the possibility of inflated Type I error rates must be taken into account when interpreting these findings. Consequently, findings should be interpreted with caution, and future investigations should aim for larger sample sizes to enhance the reliability and robustness of their conclusions and employ strategies to manage the risk of Type I errors and variance inflation.

The 8-minute exposure to the TV programs, while aligned with durations used in some previous studies, may have been too brief to produce significant or lasting cognitive effects, particularly on relatively stable functions such as visual attention or more complex components of inhibitory control. Additionally, the study did not analyze other potentially influential features of the selected programs or the specific cognitive demands embedded in the content beyond the broad "fantastical" versus "realistic" categorization. Future research would benefit from exploring a range of exposure durations and conducting more detailed analyses of content characteristics, such as narrative or language complexity, to better understand their specific cognitive impact.

On the other hand, our assessment of visual attention was confined to the gap task within the prosaccade paradigm. While this task provides specific insights into attentional disengagement and saccadic reaction time (bottom-up attention), it offers only a partial view of the multidimensional nature of visual attention. It may not capture other critical aspects such as sustained attention, selective attention, attentional control (top-down attention), or attentional flexibility, which could be differentially affected by fantastical or realistic TV content. More broadly, while this study focused on visual attention and inhibitory control, future research would benefit from employing a wider range of eye-tracking EF tasks. This would allow for a more comprehensive understanding of how different media content might impact various components of the EF system (e.g., working memory, planning, cognitive flexibility).

Finally, another limitation is the lack of explicit control for, or detailed reporting on, participants’ habitual media use and prior exposure to the types of content presented. Individual differences in media consumption patterns may significantly influence baseline EF abilities as well as children's sensitivity to media-based stimuli. Future studies should collect and analyze data on participants’ media usage to control for this potential confound and provide more nuanced interpretations of media effects.

Conclusion

This study investigated the impact of fantastical and realistic TV content on visual attention and inhibitory control in children with and without RD. While initial findings did not support the hypothesis of significant differences in visual attention across groups or conditions, analyses revealed nuanced interactions between age, behavioral/emotional problems, and TV exposure. Regarding inhibitory control, the study found varying effects across components, with significant decreases in accuracy observed in both fantastical RD and realistic TD groups. Furthermore, age and behavioral/emotional problems significantly influenced inhibitory control performance across groups. This observed vulnerabilities in children with RD, particularly in relation to hyperactivity and emotional problems, suggest the need for cautious media guidance. For instance, educators and parents may consider moderating the use of highly fantastical programs, especially before cognitively demanding activities such as learning or reading tasks. For children with RD, selecting media content with simpler narratives and fewer fantastical elements could help reduce potential cognitive load and support better engagement with subsequent tasks that rely on EFs. In conclusion, these findings highlight the potential complex interplay between individual characteristics, TV content, and cognitive development, emphasizing the need for further research to understand the short and long-term implications of media exposure for children with diverse needs.

Abbreviations

RD

Reading difficulties

TD

Typically developing

EF

Executive function

TV

Television

SDQ

Strengths and Difficulties Questionnaire

Authors' contributions

ZB, SA, and SS equally contributed to all aspects of the manuscript and reviewed and approved the final version.

Funding

No specific funding was received for this research.

Data availability

The data supporting this study’s findings are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Informed consent was obtained from the parents of all participating children. All procedures adhered to the ethical standards of the responsible committee on human research (Ethics Committee of the Research Department, Shahid Beheshti University, Tehran, Iran; approval code: IR.SBU.REC.1402.101).

Consent of publication

Not applicable.

Competing interests

The authors declare no competing interests.