Manual Dexterity is not Related to Media Viewing but is Related to Perceptual Bias in School-Age Children

Although the media can have both negative and positive effects on children’s cognitive and motor functions, its influence on their perceptual bias and manual dexterity is unclear. Thus, we investigated the association between media viewing time, media preference level, perceptual bias, and manual dexterity in 100 school-aged children. Questionnaires completed by children and their parents were used to ascertain media viewing time and preference levels. Perceptual bias and manual dexterity were measured using the visual-tactile temporal order judgment task and Movement Assessment Battery for Children—2nd edition, respectively. There were significant positive correlations between age and media viewing time and between media viewing time and media preference level. There was also a significant negative correlation between visual bias and manual dexterity. Hierarchical multiple regression analysis revealed that increasing visual bias was a significant predictor of decreasing manual dexterity. Further, children with low manual dexterity showed significant visual bias compared to those with high manual dexterity, when matched for age and gender. The present results demonstrated that, in school-aged children, although viewing media was not associated with perceptual bias and manual dexterity, there was a significant association between perceptual bias and manual dexterity.


Introduction
Studies have suggested that media can have both positive and negative effects on health, cognitive abilities, and motor function in children [1][2][3][4]. The term media includes broadcast media, such as TVs and movies, and interactive media, such as social media and video games, which allow users to consume and create content [1][2][3][4]. The former is passive media, while the latter is active media.
In children, time spent watching TV, DVDs/videos, and internet content is associated with poor health conditions such as obesity [5][6][7][8][9][10] and sleep disorders [11][12][13][14][15]. In addition, media viewing time has been shown to have a negative effect on cognitive functions such as delayed language such as a touch screen tablet on children's manual dexterity [26,[35][36][37], the effects of general media viewing, including passive media, on manual dexterity have not been clarified in children. In addition, an increase in visual bias has been shown to lead to a decrease in manual dexterity in children with motor impairment [45,46]. However, the relationship between visual bias and manual dexterity in children with typical development has not been clarified. Frequent exposure of media may increase perceived visual bias and, consequently, decrease perceived tactile bias. In addition, considering the important relationship between tactile sensation, visual bias, and manual dexterity shown by previous studies [41][42][43][44][45][46], an increase in visual bias may lead to a decrease in manual dexterity, even in children with typical development.
Therefore, an increase in media viewing in children may increase attention and sensitivity to visual information and increase their bias towards visual information, resulting in a decrease in manual dexterity. To test this hypothesis, we conducted questionnaires on media viewing and measured perceptual bias and manual dexterity of school-aged children. The questionnaires on media viewing included the average viewing time per day and preference level for media in children. We also used the visuo-tactile temporal order judgment task as a quantitative measure of perceptual bias in the current study. The temporal order judgment paradigm enabled characterization of biases indicated by the shift of the judgment to the advantage of one of the two stimuli and quantitative measurement of the effects of perceptual bias [40,47,48]. We used the manual dexterity test of the Movement Assessment Battery for Children-2nd edition [49], which is an international standard evaluation battery of motor function in children, to measure manual dexterity in the current cohort. Correlation and multiple regression analyses were performed to examine the relationships between these variables.

Participants
A total of 100 school-aged children (mean age ± standard deviation (SD), 9.2 ± 1.9 years; range, 6-12 years; 40 male participants; 87 right-handed) enrolled in regular classes at public primary schools participated in the current study. Children with typical development who were enrolled in regular classes at public primary schools in Osaka and Nara, Japan, were recruited for this study. The exclusion criteria consisted of the following: (1) a general medical condition (e.g., cerebral palsy, hemiplegia, and muscular dystrophy), (2) diagnosis of a developmental disorder (e.g., autism spectrum disorder, attention deficit hyperactivity disorder, developmental coordination disorder, learning disorder, and pervasive developmental disorder), or (3) diagnosis of intellectual disability. Eligibility was confirmed by interviewing parents and the results of regular checkups, which were provided by the school doctor at each school.
All experimental procedures were approved by the local ethics committee of the Graduate School and Faculty of Health Sciences at Kio University (approval number: R1-22). There were no foreseeable risks, and no personally identifying information was collected. The participants (children and their parents) provided background information and written informed consent. The procedures complied with the ethical standards of the 1964 Declaration of Helsinki regarding the treatment of human participants in research. Table 1 summarizes the participants' age, sex, and preferred hand distributions.  6  8  3  5  8  0  7  16  6  10  13  3  8  17  5  12  12  5  9  14  7  7  11  3  10  16  7  9  14  2  11  12  6  6  12  0  12  17  6  11  17  0   Total  100  40  60 87 13 n = number.

Procedures
All participating children completed the media viewing questionnaire, visual-tactile temporal order judgment (TOJ) task, and manual dexterity test of the Movement Assessment Battery for Children-2nd edition (M-ABC-2). The questionnaire and two experimental tasks were performed in random order for each child. The questionnaire, TOJ task, and manual dexterity test took less than 10, 20, and 30 minutes to complete, respectively. Therefore, each child completed all trials within 1 hour.

Questionnaire on Media Viewing
The children and their parents answered the questionnaire on media viewing. Media here included TV, DVD, internet content, such as YouTube and Netflix, and various video games. Media provision equipment included TVs, PCs, tablets, smartphones, and game devices such as a PlayStation and Nintendo Switch. There were two questions in the media viewing questionnaire. Question 1: The average time per day for children's media viewing (as an average for weekdays, weekends, and holidays). This question did not include parents' media viewing time and was answered as a single answer. Possible answers included the following: 0, 1, 2, 3, 4, and 5 hours or more. These answers were converted to 0, 1, 2, 3, 4, and 5 points, respectively [50][51][52]. Question 2: Preferred media level, which was answered by the children alone. Media preference level was measured using the following 7-item scale: 3, I like the media very much.; 2, I like the media.; 1, I like the media a bit.; 0, I do not like or hate the media (It is neither).; −1, I hate the media a bit.; −2, I hate the media.; −3, I hate the media very much.

Temporal Order Judgment Task
Perceptual biases were measured using the TOJ task [53][54][55][56][57] (Figure 1). Two stimuli (visual-flash and tactile-vibration) were presented in various stimulus onset asynchronies (SOAs). The child then had to determine which stimulus (visual or tactile) was presented first. The child performed this visuo-tactile TOJ task with a TOJ task device (Keio method, UT-0021, Medical Try System, Tokyo, Japan). The visual stimulus was presented as a green LED on the LED panel (UT-0021-2, Medical Try System, Tokyo, Japan). The luminance of the visual stimulus was 40 cd/m 2 , and the duration of the visual stimulus was 1 ms. A 1-ms tactile stimulus (converted to vibration by pneumatic pressure) controlled by a 1-V signal from the vibration box (UT-0021-1, Medical Try System, Tokyo, Japan) was presented to the preferred index finger.

Figure 1.
Temporal order judgment (TOJ) task. The task was presented to the child with a visuo-tactile TOJ device (Keio method, UT-0021, Medical Try System, Tokyo, Japan). This device includes an LED panel (UT-0021-2, Medical Try System, Tokyo, Japan) and vibration box (UT-0021-1, Medical Try System, Tokyo, Japan), which provided the visual and tactile stimuli, respectively. The child put the index finger of their preferred hand in the hole of the vibration box and touched the vibrotactile stimulator. Therefore, the child could not observe the tactile stimulus. The child was instructed to watch the LED panel. Conditions were set in the TOJ task device so that the visual stimulus was presented 0, 50 or 100 ms earlier than the tactile stimulus, or vice versa. A blackout curtain prevented the child from seeing outside the experimental chamber.
The stimulation condition included the following five conditions: at −100, −50, 0, 50, 100 ms, i.e., tactile stimulation was administered 50 or 100 ms earlier than the other in the −100 and −50 ms conditions, while visual stimulation was administered first in the 50 and 100 ms conditions; visual and tactile stimulation was synchronous in the 0 ms condition. In the actual test, the five stimulation conditions were considered a set, and each child performed five sets; the trial order was randomized. Therefore, each child completed 25 trials in total. Before starting the TOJ task, simple stimulus tests were used to confirm that children had no problems with vision and touch. Specifically, the tactile and visual stimuli used for the TOJ task were administered five times alone in the absence of the other stimuli to ascertain whether all the children were able to perceive the tactile and visual stimuli. The TOJ task was conducted after sufficient explanation and practice were provided to the child.
The "visual first" response probability for each of several SOA conditions (−100, −50, 0, 50, 100) was then calculated for the TOJ task. The following formula was used to fit the logistic curves to the "visual first" response probability in the TOJ task [58-60]:

exp
where t is the SOA, P(t) is the probability of the "visual first" response, a is the steepness of the fitted curve, and tPSE is the observer's point of subjective equality (PSE), the last of which demonstrates the SOA where "visual first" and "tactile first" judgment probabilities are equal (50%). A nonlinear least squares algorithm was used to fit the data in MATLAB R2014b (MathWorks, MA, USA). The PSE of each child is a quantitative indicator of perceptual bias of the individual; a large negative PSE value indicates that visual bias is strong, and a large positive PSE value indicates that tactile bias is strong. Therefore, a PSE value approaching 0 ms demonstrated no biased perception. Temporal order judgment (TOJ) task. The task was presented to the child with a visuo-tactile TOJ device (Keio method, UT-0021, Medical Try System, Tokyo, Japan). This device includes an LED panel (UT-0021-2, Medical Try System, Tokyo, Japan) and vibration box (UT-0021-1, Medical Try System, Tokyo, Japan), which provided the visual and tactile stimuli, respectively. The child put the index finger of their preferred hand in the hole of the vibration box and touched the vibrotactile stimulator. Therefore, the child could not observe the tactile stimulus. The child was instructed to watch the LED panel. Conditions were set in the TOJ task device so that the visual stimulus was presented 0, 50 or 100 ms earlier than the tactile stimulus, or vice versa. A blackout curtain prevented the child from seeing outside the experimental chamber.
The stimulation condition included the following five conditions: at −100, −50, 0, 50, 100 ms, i.e., tactile stimulation was administered 50 or 100 ms earlier than the other in the −100 and −50 ms conditions, while visual stimulation was administered first in the 50 and 100 ms conditions; visual and tactile stimulation was synchronous in the 0 ms condition. In the actual test, the five stimulation conditions were considered a set, and each child performed five sets; the trial order was randomized. Therefore, each child completed 25 trials in total. Before starting the TOJ task, simple stimulus tests were used to confirm that children had no problems with vision and touch. Specifically, the tactile and visual stimuli used for the TOJ task were administered five times alone in the absence of the other stimuli to ascertain whether all the children were able to perceive the tactile and visual stimuli. The TOJ task was conducted after sufficient explanation and practice were provided to the child.
The "visual first" response probability for each of several SOA conditions (−100, −50, 0, 50, 100) was then calculated for the TOJ task. The following formula was used to fit the logistic curves to the "visual first" response probability in the TOJ task [58][59][60]: where t is the SOA, P(t) is the probability of the "visual first" response, a is the steepness of the fitted curve, and tPSE is the observer's point of subjective equality (PSE), the last of which demonstrates the SOA where "visual first" and "tactile first" judgment probabilities are equal (50%). A nonlinear least squares algorithm was used to fit the data in MATLAB R2014b (MathWorks, MA, USA). The PSE of each child is a quantitative indicator of perceptual bias of the individual; a large negative PSE value indicates that visual bias is strong, and a large positive PSE value indicates that tactile bias is strong. Therefore, a PSE value approaching 0 ms demonstrated no biased perception.

Manual Dexterity Test of the Movement Assessment Battery for Children-2nd Edition (M-ABC-2)
The manual dexterity test of the M-ABC-2 [49] is a standardized, age-adjusted test to identify motor problems in children, in which different tasks are administered to children in different age bands. The M-ABC-2 has good test retest reliability (minimum value at any age was 0.75), inter-rater value (0.70), and concurrent validity [49]. This test has the following three age bands: 3-6, 7-10, and 11-16 years.
Our study included children aged 6 to 12 years of age. Each child took three tests that were appropriate for their age band. The 6-year-old children were in age band 1 and were administered the following three tests: posting coins test, threading beads test, and drawing trail I test. The children aged 7-10 years old were in age band 2 and were administered the following tests: placing pegs test, threading lace test, and drawing trail II test. The children aged 11-12 years old were in age band 3 and were administered the following tests: turning pegs test, triangle with nuts and bolts test, and drawing trail III test. According to the examiner's manual of M-ABC-2, standard scores of the participants are calculated from the obtained raw scores. The standard score reflects the degree of manual dexterity for each year of age, in which a higher standard score represents improvement of manual dexterity within each age group. A specifically trained and certified physical therapist administered all of these assessments.

Statistical Analyses
All statistical analyses were performed using SPSS ver. 26 (SPSS, Chicago, IL, USA). Age, media viewing time, media preference level, perceptual bias, and manual dexterity were analyzed using correlation and hierarchical multiple regression analyses. In addition, between-group comparisons based on manual dexterity scores were also performed.

Correlation Analysis
Since age, media viewing time, media preference level, perceptual bias, and manual dexterity data were not normally distributed, data were analyzed using a Spearman's correlation coefficient by rank test. The level of statistical significance was set at p < 0.05.

Hierarchical Multiple Regression Analysis
In the current study, we hypothesized that an increase in media viewing would increase visual bias and eventually reduce manual dexterity. Therefore, we performed hierarchical multiple regression analysis (forced entry method) with manual dexterity as the dependent variable and age, media viewing time, media preference level, perceptual bias, and interaction terms as independent variables. We created three interaction terms, considering the possibility that the media viewing time and the media preference level have become moderator variables, which may enhance the relationship between the increase in visual bias and the decrease in manual dexterity. These were interaction term-1 (perceptual bias × media viewing time), interaction term-2 (perceptual bias × media viewing level), and interaction term-3 (perceptual bias × media viewing time × media preference level). The interaction terms were calculated by multiplying the values obtained by centering each variable, taking into account the multicollinearity. In model 1, age, media viewing time, media preference level, and perceptual bias were independent variables. In model 2, interaction term-1 (perceptual bias × media viewing time) was added to model 1 as an independent variable. In model 3, interaction term-2 (perceptual bias × media preference level) was added to model 2 as an independent variable. In model 4, interaction term-3 (perceptual bias × media viewing time × media preference level) was added to model 3 as an independent variable. The statistical significance level was set at p < 0.05.

Comparison between Groups
Since children in the 63rd percentile (standard manual dexterity test score, 11) or higher have relatively greater manual dexterity ability, they were classified as the 'high manual dexterity group'. In contrast, children in the 50th percentile (standard manual dexterity score, 10) or lower were classified as the 'low manual dexterity group'. Gender and preferred hand between the two groups were compared using a chi-squared test for independence. Data (age, media viewing time, media preference level, and perceptual biases) of both groups were compared using a Mann-Whitney U test since data were not normally distributed. The level of statistical significance was set at p < 0.05. The effect size was also calculated [61]. Table 2 shows a summary of all data obtained from 100 children. The raw data of all participants are shown in Table A1.  Table 3 outlines the results of the correlation analysis. There was a significant correlation between age (years) and media viewing time (hour, Questionnaire 1) (p = 0.003, r = 0.293) ( Table 3). There was also a significant correlation between media viewing time (hour, Questionnaire 1) and media preference level (Questionnaire 2) (p = 0.007, r = 0.269) ( Table 3). Further, there was a significant correlation between perceptual biases (ms, PSE) and manual dexterity (standard score) (p < 0.001, r = 0.537; Table 3, Figure 2). However, there was no significant correlations between age, media viewing time, and media preference level and perceptual biases or manual dexterity (Table 3).  Correlation between perceptual bias and manual dexterity. Scatter plot showing the relationship between the temporal order judgment task and manual dexterity test. The vertical axis shows the results of the manual dexterity test (standard score) of each child. An increase in the standard score indicates an improvement in manual dexterity. The horizontal axis shows the results of the temporal order judgment task (PSE) for each child. Negative PSE values on the horizontal axis represent "visual first" indicators that were higher although the tactile stimulus was actually faster than the visual stimulus. Thus, an increase in the negative PSE value represents an increase in visual bias. Conversely, positive PSE values on the horizontal axis represent the "visual first" indicators when the visual stimulus was actually earlier than the tactile stimulus. Thus, an increase in the positive PSE value represents an increase in tactile bias. Therefore, the closer the PSE value is to 0 (middle line), the lower the perception bias. PSE, point of subjective equality.

Hierarchical Multiple Regression Analysis Results
The results of the hierarchical multiple regression analysis are summarized in Table 4. The results of the hierarchical multiple regression analysis showed that model 1 had the best fit compared to other models (model 2-4). For manual dexterity, compared to the other models (model 2-4), model 1 showed the highest coefficient of determination adjusted for the degrees of freedom (adjusted R2), a significant change in the multiple coefficient of determination (R2), and the lowest Akaike information criterion (AIC) and Bayesian information criterion (BIC), indicating that model 1 was the best fit for the data. In addition, there was no interaction effect between perceptual bias and media viewing/preference level. Therefore, the media viewing time/preference level was not a moderator variable that strengthened the relationship between increasing visual bias and decreasing manual dexterity. Figure 2. Correlation between perceptual bias and manual dexterity. Scatter plot showing the relationship between the temporal order judgment task and manual dexterity test. The vertical axis shows the results of the manual dexterity test (standard score) of each child. An increase in the standard score indicates an improvement in manual dexterity. The horizontal axis shows the results of the temporal order judgment task (PSE) for each child. Negative PSE values on the horizontal axis represent "visual first" indicators that were higher although the tactile stimulus was actually faster than the visual stimulus. Thus, an increase in the negative PSE value represents an increase in visual bias. Conversely, positive PSE values on the horizontal axis represent the "visual first" indicators when the visual stimulus was actually earlier than the tactile stimulus. Thus, an increase in the positive PSE value represents an increase in tactile bias. Therefore, the closer the PSE value is to 0 (middle line), the lower the perception bias. PSE, point of subjective equality.

Hierarchical Multiple Regression Analysis Results
The results of the hierarchical multiple regression analysis are summarized in Table 4. The results of the hierarchical multiple regression analysis showed that model 1 had the best fit compared to other models (model 2-4). For manual dexterity, compared to the other models (model 2-4), model 1 showed the highest coefficient of determination adjusted for the degrees of freedom (adjusted R2), a significant change in the multiple coefficient of determination (R2), and the lowest Akaike information criterion (AIC) and Bayesian information criterion (BIC), indicating that model 1 was the best fit for the data. In addition, there was no interaction effect between perceptual bias and media viewing/preference level. Therefore, the media viewing time/preference level was not a moderator variable that strengthened the relationship between increasing visual bias and decreasing manual dexterity.  The detailed results of the hierarchical multiple regression analysis (model 1) are as follows: with manual dexterity as a dependent variable, only perceptual bias was a significant independent variable (β = 0.501, p < 0.001). Age, media viewing time, and media preference level were not significant independent variables of manual dexterity. The relationship between manual dexterity and perceptual bias could be modeled with the following equation: Manual dexterity = 13.080 + (0.033 × perceptual bias), resulting in the following results: R = 0.503, R 2 = 0.253, Adjusted R 2 = 0.221, p < 0.001. In addition, there was no multicollinearity effect (Table 4).

Comparison Results between Groups
As a result of grouping based on the manual dexterity test score, 71 children (25 male participants; 63 right-handed) and 29 children (15 male participants; 24 right-handed) were grouped into high and low manual dexterity groups, respectively. Table 5 summarizes the data for each group. There was no significant difference in sex (χ 2 (0.95) = 3.841, χ 2 = 2.339, p = 0.126) and preferred hand (χ 2 (0.95) = 3.841, χ 2 = 0.650, p = 0.420) between groups. Further, there were no significant differences in age (z = −0.284, p = 0.776, effect size (r) = −0.03), media viewing time (z = −0.474, p = 0.635, effect size (r) = −0.05), and media preference level (z = −0.051, p = 0.959, effect size (r) = −0.01) among the groups. Figure 3A shows the "visual first" response probability curves for both groups, and Figure 3B shows the perceptual biases (PSE) comparison results for both groups. The PSE of the low manual dexterity group (mean ± standard deviation, −43.4 ± 44.0) was significantly lower than the PSE in the high manual dexterity group (mean ± standard deviation, −4.3 ± 33.8) (z = −4.543, p < 0.001, effect size (r) = −0.45; Figure 3B). This indicated that the low manual dexterity group had a significantly stronger visual bias than the high manual dexterity group.   Figure 3A shows the "visual first" response probability curves for both groups, and Figure 3B shows the perceptual biases (PSE) comparison results for both groups. The PSE of the low manual dexterity group (mean ± standard deviation, −43.4 ± 44.0) was significantly lower than the PSE in the high manual dexterity group (mean ± standard deviation, −4.3 ± 33.8) (z = −4.543, p < 0.001, effect size (r) = −0.45; Figure 3B). This indicated that the low manual dexterity group had a significantly stronger visual bias than the high manual dexterity group.

Discussion
The current study analyzed the association between media viewing time, media preference level, perceptual bias, and manual dexterity in school-aged children (6-12 years). There were significant correlations between age and media viewing time and between media viewing time and media

Discussion
The current study analyzed the association between media viewing time, media preference level, perceptual bias, and manual dexterity in school-aged children (6-12 years). There were significant correlations between age and media viewing time and between media viewing time and media preference level. This indicated that children growing older is associated with an extension in media viewing time, which in turn is related to an increased preference for media. However, there was no significant correlation between media viewing time and/or media preference level and perceptual bias and/or manual dexterity. Conversely, there was a significant correlation between the increase in visual perceptual bias and the decrease in manual dexterity. Furthermore, the results of hierarchical multiple regression analysis showed that among the variables (i.e., age, media viewing time, media preference level, and perceptual bias), only perceptual bias was a significant independent variable for manual dexterity. Media viewing time and media preference level did not strengthen the relationship between increasing visual bias and decreasing manual dexterity. In addition, children with low manual dexterity showed significant visual bias compared to those with high manual dexterity.
The significant correlation between age and media viewing time is consistent with previous studies, which have reported this association in children aged <3 years and 3-16 years old [35,51,[62][63][64][65][66][67][68]. The current study also demonstrated a significant correlation between media viewing time and media preference level.
However, there was no significant correlations between media viewing time and/or media preference level and perceptual bias and/or manual dexterity in the current study. In addition, hierarchical multiple regression analysis did not reveal the effect of media viewing on perceptual bias and manual dexterity. Previously, various adverse effects of media viewing on cognitive functions and motor functions in children, such as language development delay [16,17], poor health [9,22], and attention problems [18], have been reported. However, these effects may be mitigated by factors such as parenting style [34], type of content and program [69], and co-viewing with a parent [70]. This could be attributed to the variation in parenting styles, i.e., authoritative or permissive parenting based on Baumrind's conceptualization [71], where the style of parenting determines the degree of discipline and the children's responsiveness. Further, educational internet content and TV programs can have a positive impact on cognitive development in children [4]. In addition, socioeconomic status, such as education, income and deprivation, is associated with children's media viewing and health status. Many studies have found an increase in media viewing time and sedentary time as the socioeconomic status decreases [72][73][74][75][76][77][78]. An increase in sedentary time can lead to a decrease in the time spent engaging in play and sports activities, which in turn can reduce manual dexterity. However, since we did not collect data on parenting style, type of content and program, and socioeconomic status, the effects of these factors in the current study are completely speculative.
There is also some evidence that active video game enhances cognitive functions, such as visual processing and attention, and contributes to the improvement of motor control [79]. Thus, despite the negative risks, media can also have positive effects on children. In addition to video games that required movement, media included in the current study also included TV and DVDs/videos, which only required passive viewing. Therefore, our findings may have been different if media were classified into interactive and broadcast media, which are active and passive, respectively. This is also supported several studies investigating the relationship between the use of a touch screen tablet, which can be manipulated and drawn on by touching the screen, and manual dexterity. Lin et al. [26] revealed that manual dexterity improves in children who did not use tablets compared with those who did. This was attributed to the fact that real actions of grasping, drawing, and manipulating objects required more muscle power, coordination, and dexterity than the virtual actions on a tablet [26]. However, other studies have reported discrepant results to Lin et al. [26]. Several previous studies demonstrated that the use of tablets is significantly associated with an improvement in manual dexterity, emphasizing the similarities between the virtual environment on the tablet screen and the real physical environment [35][36][37]. Therefore, the type of media used also seems to affect manual dexterity. Since we did not differentiate the type of media in the current study, it is not easy to discern whether this would have affected perceptual bias/manual dexterity. However, the relationship between media viewing time/media preference level and perceptual bias/manual dexterity should be explored in further future studies considering the limitations of the current study.
Importantly, the current study demonstrated a significant correlation between the increase in visual bias and decrease in manual dexterity in school-aged children with typical development. Further, increased visual bias was a significant predictor of reduced manual dexterity. In addition, participants with low manual dexterity showed significant visual bias compared to those with high manual dexterity. It is important to note that the children in the current study had typical development and had neither visual nor tactile problems when completing the simple stimulus tests prior to the TOJ task. Therefore, the current results could not be attributed to visual or tactile disorders. The tactile sensation of the hand is a prerequisite for manual dexterity, such as grasping objects, manipulating objects, and handwriting [41][42][43][44]. A study demonstrated that adult string players have an expanded cortical representation of somatosensory sensations of the fingers [80], which is direct evidence of the important relationship between manual dexterity and somatosensory sensation. Further, preschool children with motor delays show a significant correlation between tactile sense and fine motor skills [81]. Further, children with cerebral palsy [82] and Duchenne muscular dystrophy [83] show a significant relationship between hand tactile function and manual dexterity. Previous studies have also demonstrated that children with clumsy movements reported relying on visual information, and not tactile information, for hand movements [45,46]. Several previous studies have suggested that prioritizing visual information, not tactile information that is important for movement, adversely affects the success of the movement task [45,46,[84][85][86][87][88][89]. Both visual and tactile information are important resources for manual dexterity. However, even if there is no visual information, manual dexterity can be completed with tactile information only. Therefore, the present results showed that prioritizing visual information over tactile information when visual and tactile stimuli are input almost simultaneously leads to poor manual dexterity, even in school-aged children with typical development.

Limitations of the Current Study and Future Directions
Since the current study did not investigate the type of media (i.e., broadcast media or interactive media, active media or passive media, etc.), type of content (educational or otherwise), parenting style or socioeconomic status, we cannot determine whether these factors influenced the results in the current study. These factors play an important role in understanding the effect of media viewing on children's cognitive and motor functions. Thus, further research, which includes measures of these factors, is required to better understand the relationship between media viewing and perception bias/manual dexterity in children.
All children who participated in the current study could detect a 1-ms tactile stimulus controlled by a 1-V signal in the TOJ task. However, the children's tactile threshold (tactile sensitivity) was not measured. Therefore, differences in tactile sensitivity may have affected the results of the manual dexterity test. Taking this limitation into account, future studies will also require measurement of tactile sensitivity.

Conclusions
There was no significant correlation between media viewing time and/or media preference level, and perceptual bias and/or manual dexterity in school-aged children with typical development. However, there was a significant association between age and media viewing time and between media viewing time and media preference level in the current cohort. Further, there was also a significant correlation between visual bias and manual dexterity. Finally, increased visual bias was a significant predictor of reduced manual dexterity, which indicated that children with relatively low manual dexterity had strong visual bias.

Acknowledgments:
The authors wish to acknowledge and thank all the children and their parents who participated in this study.

Conflicts of Interest:
The authors declare no conflict of interest.