Abstract
Children's internalising and externalising behaviours have been widely studied due to their implications for socioemotional development. Although media use has been identified as a relevant contextual factor, growing evidence suggests that parenting practices play a critical role in shaping how media engagement relates to behavioural outcomes. This study examines the associations between parental media monitoring strategies and internalising and externalising behaviours in a community sample of 411 Portuguese middle-school children aged 10 to 12 years. Data were collected using child- and parent-reported questionnaires. Multiple regression analyses revealed that inconsistent parental monitoring significantly influenced both externalising and internalising behaviours across informants, even after controlling for demographic factors. Sex and socioeconomic indicators were also associated with behavioural outcomes. These findings highlight the central role of consistency in parental media monitoring and suggest that the quality of parental supervision is crucial for understanding children's behavioural adjustment in digital contexts.
Introduction
Over the past decade, children and youth have experienced a significant increase in media exposure, now averaging nearly 8 hours of daily media use, far exceeding the 2-hour limit recommended by the American Academy of Pediatrics (American Academy of Pediatrics, 2016; Rideout & Robb, 2019). While prior research has acknowledged potential benefits of media engagement, such as the promotion of prosocial behaviours (Coyne et al., 2018; Padilla-Walker et al. 2011), a growing body of literature has linked excessive and unregulated media consumption to negative developmental outcomes, including academic underachievement, behavioural problems, and substance use (Nesi et al., 2018; Twenge & Campbell, 2018). Regardless of the specific outcome, previous studies demonstrated that the influence of media on children and youth is not uniform and is shaped by familial factors, among which parental involvement is particularly salient (Nathanson, 2001; Valkenburg et al., 2013). Parental media monitoring strategies, such as restrictive, active, or inconsistent supervision play a crucial role in shaping the relationship between media exposure and the development of internalising and externalising behaviours (Gentile et al., 2012; Livingstone & Helsper, 2008). The relevance of re-examining these parental processes lies in the shifting landscape of digital childhoods. As digital access becomes normative, traditional “gatekeeping” strategies centred on total prohibition are increasingly considered obsolete and potentially counterproductive (Coyne et al., 2023a; Livingstone & Blum-Ross, 2020).
This study moves beyond the binary of access versus non-access to media devices by demonstrating that the quality and consistency of monitoring emerge as a key correlate of behavioural health (Sanders et al., 2022). Specifically, using a sample of Portuguese middle-school children, the current study seeks to explore the associations between parental media monitoring strategies and children's internalising and externalising behaviours. Furthermore, the study explores whether parental consistency provides a stable regulatory context associated with children's behavioural adjustment. While inconsistent parental behaviours may undermine a child's sense of predictability and emotional security, consistent monitoring provides the necessary scaffolding for children to internalise behavioural norms and develop the self-regulatory skills needed to manage digital stimuli (Pinquart, 2021; Sanders et al., 2022).
Media use and child development: Theoretical frameworks
The concept of media use has been widely employed in recent research, although its definition varies across studies. Some authors adopt relatively narrow operationalisations focused on activities such as television viewing or streaming platforms (Hill et al., 2016), whereas others employ broader definitions that include computers, mobile devices, social media, and internet use (Cain & Gradisar, 2010). Following Padilla-Walker's et al. (2020) definition, in the present study media use is conceptualised as a multifaceted set of screen-based activities, including television viewing, video games, social media use, internet browsing, messaging, and media multitasking. This perspective reflects the increasingly complex and interactive nature of children's digital environments (Nesi et al., 2018; Valkenburg & Piotrowski, 2017).
Increasingly, research suggests that the developmental impact of media depends not only on the amount of exposure, but also on the context in which media use occurs (Clemente-Suárez et al., 2024). Accordingly, the present study does not conceptualise media use as a primary explanatory variable, but rather as a contextual condition within which parental monitoring practices operate.
From a theoretical perspective, several frameworks explain the relationship between media and behaviour. For example, the General Aggression Model (Anderson & Bushman, 2002) and Social Learning Theory (Bandura, 1977) suggest that repeated exposure to aggressive or maladaptive media content may contribute to the acquisition and reinforcement of behavioural scripts associated with aggression and poor self-regulation, particularly within contemporary digital environments (Coyne et al., 2023a). These approaches have been particularly influential in explaining externalising outcomes, such as aggression and impulsivity, including how repeated exposure to violence may normalise proactive aggressive behaviours (Dodge, 1991).
However, these classical frameworks primarily focus on direct exposure effects and are less informative in explaining broader dimensions of adjustment, particularly internalising difficulties, or the role of family processes in shaping children's media experiences (Coyne et al., 2023a). To address these gaps, the current study adopts an ecological-transactional perspective (Bronfenbrenner & Morris, 2006), emphasising that children's behavioural adjustment develops within broader relational and contextual environments. Within this framework, parental monitoring is conceptualised as an important contextual factor associated with children's opportunities to develop behavioural expectations, emotion regulation, and self-regulatory skills (Compas et al., 2017; Eisenberg et al., 2010).
In this context, parenting practices characterised by structure and predictability may support children's behavioural adjustment, whereas inconsistent parental monitoring, characterised by unpredictable or erratic supervision practices, may be associated with greater emotional and behavioural difficulties (Pinquart, 2021; Sanders et al., 2022). Accordingly, the present study examines whether different parental media monitoring strategies are differentially associated with children's internalising and externalising behaviours within contemporary digital contexts.
Internalising and externalising behaviours in the context of media use
Problem behaviours in childhood are widely recognised as significant predictors of later maladjustment across multiple developmental domains, including behavioural, emotional, cognitive, and social functioning (Campbell et al., 2000; Liu et al., 2011; Rinaldi & Howe, 2012). Accordingly, over the last few decades, significant attention has been devoted to the understanding of the emergence and development of such behaviours. Within this field, a central analytical distinction differentiates between externalising and internalising forms of problem behaviour (Akhter et al., 2011; Alizadeh et al., 2011; Rinaldi & Howe, 2012), whose developmental trajectories follow distinct patterns across childhood (Almeida & Santos, 2024; Almeida et al., 2025).
Externalising behaviours are typically conceptualised as outwardly directed and socially disruptive, reflecting maladaptive patterns of interaction with the surrounding environment (Georgiou & Symeou, 2018; Liu, 2004). These behaviours encompass a broad spectrum of observable actions, including hyperactivity, impulsivity, attentional deficits, and emotional dysregulation, as well as more severe manifestations such as aggression, antisocial conduct, and delinquency (Alizadeh et al., 2011; Carreras et al., 2014). Given their visibility, such behaviours are more readily identified and are often associated with immediate and long-term consequences for both the individual and the community (Georgiou & Symeou, 2018). Common expressions include defiance, verbal aggression, restlessness, and property damage (Almeida et al., 2025; Rinaldi & Howe, 2012; Santos et al., 2020). Externalising behaviours tend to peak in early childhood and gradually decline with age as regulatory and executive functions mature (Moffitt, 1993; Shaw et al., 2003).
In contrast, internalising behaviours operate primarily at the intrapersonal level and are less directly observable (Achenbach & Edelbrock, 1978). These patterns involve emotional and psychological distress, including anxiety, fear, depressive symptoms, social withdrawal, shyness, and somatisation (Alizadeh et al., 2011; Carreras et al., 2014; Georgiou & Symeou, 2018; Liu, 2004; Rinaldi & Howe, 2012). Although their external impact may be less apparent, internalising problems are associated with substantial impairments in psychological functioning and are often more difficult to detect due to their covert nature (Almeida et al., 2025; Georgiou & Symeou, 2018).
While internalising and externalising behaviours have often been examined in relation to media use, growing evidence suggests that these associations are highly contingent on family and relational contexts, particularly parental monitoring practices. Evidence suggests that exposure to violent or fast-paced media content, such as in certain video games, television programmes, or unregulated online environments, can heighten physiological arousal, desensitise viewers to aggression, and normalise antisocial behaviour, thereby contributing to externalising problems (Anderson & Bushman, 2002; Stockdale & Coyne, 2020; Gentile et al., 2012). This link appears especially pronounced during early developmental stages, when self-regulation and inhibitory control are still developing (Huesmann, 2007).
At the same time, media use patterns characterised by passive or socially isolating engagement have been associated with internalising problems. For example, prolonged exposure to idealised online portrayals of peers, repetitive passive scrolling, or solitary media engagement (e.g., binge-watching or gaming alone) has been linked to feelings of inadequacy, low self-worth, and emotional withdrawal (Keles et al., 2020; Nesi & Prinstein, 2015; Twenge et al., 2018). These associations are particularly strong among children with pre-existing vulnerabilities, such as low self-esteem or limited offline social support (Frison & Eggermont, 2016).
Importantly, the relationship between media use and behavioural outcomes is not determined solely by the amount or content consumed but also by the context in which that use occurs. Without structure or guidance, children may engage with media in ways that reinforce maladaptive coping strategies or avoidance of real-world challenges. In this sense, parental media monitoring constitutes a significant contextual factor influencing how children interact with and are impacted by digital media. The strategies that parents employ to supervise, guide, or limit their children's media use can significantly buffer, or exacerbate, the impact of media engagement on both internalising and externalising behaviours (Gentile et al., 2012; Livingstone & Helsper, 2008; Nathanson, 2001).
Parental monitoring of media use
Parental monitoring of children's media use has emerged as a central factor in explaining how digital engagement shapes behavioural outcomes. This construct refers to the strategies parents use to supervise, guide, and regulate their children's interactions with media (Nathanson, 2001; Valkenburg et al., 1999), which are commonly grouped into four primary monitoring categories, namely restrictive monitoring, active monitoring, passive co-viewing, and inconsistent monitoring.
Restrictive monitoring, which involves setting explicit limits on screen time or access to certain content, has been associated with lower levels of externalising behaviours (e.g., aggression, hyperactivity) and reduced exposure to media-related risks such as violent or sexualised content, cyberbullying, and online solicitation (Bleakley et al., 2016; Chang et al., 2015; Gentile et al., 2012). However, overly rigid restrictions, especially when imposed without explanation or negotiation, can elicit psychological reactance or covert media use among children (Coyne et al., 2018; Nathanson, 2002). Active monitoring, defined by parent-child discussion about media content and interpretation, has consistently been linked to positive developmental outcomes, including reduced aggression, improved emotion regulation, and enhanced moral reasoning (Collier et al., 2016; Padilla-Walker et al., 2020; Valkenburg et al., 1999). In contrast, co-viewing, where parents are physically present but cognitively disengaged, tends to increase children's screen time and reduce opportunities for critical media literacy (Connell et al., 2015; Padilla-Walker et al., 2010). Finally, inconsistent monitoring, marked by unpredictable rules and low communication, has been associated with both internalising and externalising problems, including anxiety, depression, and oppositional behaviour (Modecki et al., 2022).
Empirical evidence further shows that parental monitoring shapes how children engage with media and how media-related experiences relate to behavioural adjustment, either buffering or amplifying potential impacts (e.g., Collier et al., 2016; Padilla-Walker et al., 2020; Valkenburg et al., 2013). For instance, it was found that restrictive monitoring might reduce the association between screen time and aggression (Nikken & Jansz, 2014), while parental involvement was associated with less indirect effects of violent media exposure on externalising problems through its influence on normative beliefs about aggression (Gentile et al., 2012). Moreover, inconsistent or absent monitoring may amplify the negative effects of easy and early device accessibility, leading to greater behavioural difficulties and poorer academic adjustment (Lissak, 2018).
A critical but often overlooked dimension of parental mediation is the role of parental consistency. Inconsistent parental monitoring is defined not merely by a low frequency of supervision, but by the unpredictability and erratic nature of rule enforcement (Padilla-Walker et al., 2018). It is crucial to distinguish this from low involvement, which refers to a chronic absence of parental oversight. In contrast, inconsistent monitoring aligns with the broader literature on inconsistent discipline, describing a pattern where parents oscillate between permissive and restrictive responses to the same behaviour (Coyne et al., 2021a). Under this pattern, parents may strictly prohibit a digital activity on one occasion but permit it on another, often influenced by situational stressors, parental mood, or fatigue rather than a stable educational project (Coyne et al., 2021a). From a developmental perspective, this unpredictability functions as a form of intermittent reinforcement, making it difficult for the child to identify stable contingencies for their behaviour and predict the consequences of digital engagement. Consequently, this lack of structure undermines the “scaffolding” necessary for children to internalise social norms and is theorised to hinder the development of effortful control, which is a key self-regulatory component in preventing both externalising impulses and internalising distress (Crone & Konijn, 2018; Eisenberg et al., 2010; Sanders et al., 2022).
While the literature conceptually identifies four primary parental monitoring strategies (restrictive, active, co-viewing, and inconsistent monitoring), the current study focuses on three: restrictive, active, and inconsistent monitoring. This decision is justified by the internal consistency results of the measure used in this research, which was specifically designed to assess these three dimensions within digital contexts. Co-viewing was not included in the analysis as it did not emerge as a distinct factor in the validated instrument used for this specific sample, a common finding in contexts where the individualisation of digital devices reduces shared screen time (Connell et al., 2015).
The importance of using multiple informants to assess children's internalising and externalising behaviours
Over the past decades, a substantial body of research has emphasised the importance of using multiple informants, such as parents, teachers, and children, when assessing children's social-emotional development and behavioural functioning (e.g., Achenbach, 2006; Achenbach et al., 1987; De Los Reyes & Kazdin, 2005; Santos et al., 2020). Relying on a single informant tends to reflect only that individual's subjective perspective, which may limit the accuracy and comprehensiveness of the assessment. In contrast, integrating multiple sources of information allows for a more robust understanding of children's strengths and difficulties, improving the reliability and validity of the data and supporting more effective identification, classification, and intervention efforts (Achenbach et al., 1987; De Los Reyes et al., 2015). This approach is particularly relevant in developmental research, as children, especially at younger ages, may vary in their ability to accurately report on their own emotions and behaviours (Becker et al., 2004). In addition, different informants, particularly parents, observe children across distinct contexts, capturing variations in behaviour that might otherwise go unnoticed (Becker et al., 2004; De Los Reyes et al., 2013). Also, when examining internalising behaviours, which are often characterised by intrapersonal experiences, such as psychological and emotional processes, and may therefore be less readily observable to external informants, self-reports are particularly relevant (Graber, 2004).
Within this multi-informant framework, research has consistently documented low to moderate levels of agreement between informants, particularly between parents and children (Achenbach et al., 1987; De Los Reyes et al., 2015; Winsler & Wallace, 2002). Although such discrepancies might initially be viewed as a limitation, they are now widely understood as meaningful rather than merely reflecting measurement error (De Los Reyes, 2011; De Los Reyes et al., 2013). Differences in reports may arise from distinct perspectives, expectations, and experiences, as well as from the specific contexts in which behaviours are observed and the informant's characteristics (Achenbach et al., 2008; Antrop et al., 2002; Gross et al., 2004; Satake et al., 2003; Van der Ende, 1999). As such, disagreement across informants can provide valuable insight into how children's behaviours and experiences vary across settings and relationships (Santos et al., 2020).
Current study
Considering this rationale, the current study seeks to examine the associations between different parental media monitoring strategies (i.e., active, restrictive, and inconsistent monitoring) and the manifestation of internalising and externalising behaviours in children, based on parent and self-reports. By situating these processes within an ecological-transactional framework, it aims to clarify how the quality and consistency of parental supervision relate to child adjustment in the Portuguese context. Based on the theoretical background previously described, the following hypotheses were formulated: i) active parental monitoring is negatively associated with both internalising and externalising behaviours; ii) restrictive parental monitoring is negatively associated with both internalising and externalising behaviours; and iii) inconsistent parental monitoring is positively associated with both internalising and externalising behaviours.
Method
Participants
The study was developed with a community sample of 413 middle-school children and 385 of their parents, as presented in Table 1. Among the child participants, 52.1% were female (n = 211), with a mean age of 10.83 years (SD = 0.75). The parent sample was predominantly female (87.3%, n = 336) and had a mean age of 41.98 years (SD = 5.65). The majority of the respondents were the biological parents of the participating children (95.8%, n = 369). For the purposes of this study, the person answering the questionnaire, either the parent or the legal guardian, is labelled the parent.
Sample characteristics.
Note. M – Mean; SD – Standard Deviation.
Procedure
The study protocol received ethical approval from the Faculty Ethics Committee and authorisation from the School Environment Survey Monitoring System (MIME). Data collection took place in 2024. Schools were recruited based on their location (districts of Porto and Aveiro, Portugal), their public status (private schools were not considered), the grade levels offered (5th and 6th grades, aligning with the inclusion criteria defined for the study) and their availability to collaborate. The research team contacted each school principal to obtain formal authorisation to conduct the study. Of the thirteen schools contacted, five agreed to participate, while the remaining eight declined or did not respond (mainly due to curricular scheduling constraints, administrative workload, or ongoing participation in other research projects). Following approval, participating schools collaborated with the research team to identify eligible classes according to predefined inclusion criteria: students aged 10 to 12 years, enrolled in the 5th or 6th grades of middle school. Students with formally identified special educational needs were excluded. This decision, besides seeking to maintain sample homogeneity, took into consideration the self-report nature of the data collection and the fact that children with such needs might have some additional language, reading and comprehension difficulties, which could have impaired the answers provided in the questionnaire and introduced bias into the data. Prior to data collection, detailed information about the study's objectives and procedures was provided to school administrators, teachers, parents, and students. Written informed consent was obtained from all parents, and child assent was secured before participation.
Children completed paper-and-pencil self-report questionnaires during classroom hours. Trained members of the research team presented the study, clarified instructions, and ensured independent and voluntary participation. Parents also completed paper and pencil self-report questionnaires, which were delivered via the participating children and completed at home, without any further guidance or the presence of the researchers. Parents were given two weeks to complete the questionnaires and, after that, they were asked to return the forms to the teachers, who returned them to the researchers. The average time required to complete the parent survey was approximately 20 min. Out of a total of 516 eligible parents invited to participate, 382 provided informed consent and returned valid questionnaires, resulting in a participation rate of 74%.
Measures
Demographic variables
Demographic information was collected in both child- and parent-reported questionnaires. Specifically, data was gathered concerning sex (1 = male; 2 = female), age (years), relation with the child and socioeconomic status. Socioeconomic status (SES) was operationalised through two indicators: Perceived family income (Participants rated their household income on a 4-point scale: (1) “living comfortably,” (2) “living on current income,” (3) “difficult to live on current income,” and (4) “very difficult to live on current income”) and Perceived social class (Participants self-identified their social standing using five categories: (1) Low class, (2) Lower-middle class, (3) Middle class, (4) Upper-middle class, and (5) Upper class).
Externalising and internalising behaviours
Externalising and internalising behaviours were assessed using the Achenbach System of Empirically Based Assessment (ASEBA; Achenbach & Rescorla, 2001), specifically the Child Behaviour Checklist for Ages 6–18 (CBCL/6–18) and the Youth Self-Report (YSR/11–18). These standardised, empirically validated measures are widely used to identify emotional, social, and behavioural problems, as well as social competencies, in children. The CBCL was completed by the parents, and the YSR by the participating children, providing data from multiple informants.
Both instruments contain two broadband scales – Internalising and Externalising Problems – and eight empirically derived syndrome scales. The Internalising broadband scale assesses emotional difficulties through three syndrome subscales: Anxious/Depressed, Withdrawn/Depressed, and Somatic Complaints. The Externalising broadband scale assesses behavioural and conduct-related difficulties and is composed of the Rule-Breaking Behaviour and Aggressive Behaviour subscales. Additional syndrome scales (Social Problems, Thought Problems, and Attention Problems) are not included in either broadband scale but provide complementary information on behavioural adjustment. For each item, respondents rated the frequency or intensity of a given behaviour over the past six months using a three-point Likert scale: 0 (not true), 1 (somewhat or sometimes true), and 2 (very true or often true). Both the CBCL and YSR Portuguese versions were translated and validated by Fonseca et al. (1994) and Fonseca and Monteiro (1999), respectively, demonstrating satisfactory psychometric properties in Portuguese samples. Regarding the CBCL, for five subscales the internal consistency was above 0.70, but for the remaining sub-scales the reliability ranged from 0.53 for social problems to 0.66 for withdrawn (Fonseca et al., 1994). As for YSR, satisfactory internal consistency was found. Reported Cronbach's alpha coefficients were above .85 for the broadband scales of Internalising and Externalising Problems, indicating good reliability. At the syndrome level, alpha values generally ranged between .70 and .80, reflecting acceptable to good internal consistency. In the current study, both the externalising (child report α = .85; parent report α = .89) and the internalising (child report α = .89; parent report α = .87) broadband scales revealed good internal consistency. These findings are consistent with the reliability indices reported in the ASEBA technical manual (Achenbach & Rescorla, 2001).
Parental media monitoring
Parental media monitoring was assessed using the Perceived Parental Media Mediation Scale (Valkenburg et al., 2013) directed at the participating children and designed to capture parents’ strategies for supervising and guiding children's media use. The instrument comprises three theoretically and empirically grounded dimensions: restrictive monitoring (e.g., “My parents set rules about how long I can use screens”), active monitoring (e.g., “My parents talk to me about what I can watch or play”) and inconsistent monitoring (e.g., “Sometimes my parents allow me to use media even when they previously said no”). This multidimensional structure allows for a nuanced understanding of how parents engage with and regulate their children's media environments. In the present sample, internal consistency coefficients indicated acceptable reliability across subscales, with α = .79 for restrictive monitoring, α = .65 for active monitoring, and α = .68 for inconsistent monitoring.
Data analysis
Data were analysed using IBM SPSS Statistics version 29. Descriptive statistics and reliability analyses were first conducted to characterise the sample and examine the internal consistency of the study measures. Pearson's correlation coefficients (r) were used to explore bivariate associations among parental media monitoring strategies and internalising and externalising behaviours, based on both child- and parent-reported data. Multiple linear regression analyses were conducted to examine variables associated with internalising and externalising behaviours. Demographic variables (sex, age, and socioeconomic indicators) were entered as control variables, followed by parental media monitoring strategies.
Results
Preliminary analyses
Preliminary analyses examined sex differences across the dependent variables studied. According to child reports, boys presented significantly higher levels of externalising behaviours t(2.57), p = .01, whereas girls reported higher levels of internalising symptoms (t(2.68), p < .01). In parent reports, boys were also described as presenting significantly higher levels of externalising behaviours compared to girls (t(2.87), p < .01) although no significant sex differences emerged for internalising symptoms. These findings are consistent with the ones from previous studies, thus supporting the inclusion of sex as a control variable in subsequent multiple linear regression analyses.
Correlations between externalising behaviours, internalising behaviours and monitoring strategies
Table 2 presents the Pearson correlation coefficients among the main study variables, analysed separately according to child- and parent-reported behavioural outcomes.
Correlations among study variables
**p < .01*p < .05
Regarding child-reported measures, inconsistent parental monitoring was positively associated with both externalising behaviours (r = .19, p < .01) and internalising behaviours (r = .15, p < .01), suggesting that higher levels of unpredictable or erratic supervision are related to greater emotional and behavioural difficulties as perceived by children. Restrictive monitoring showed a weak negative association with internalising behaviours (r = –.20, p < .05), whereas no significant association was found with externalising behaviours. Active parental monitoring was not significantly associated with either child-reported externalising or internalising behaviours. In addition, restrictive and active monitoring were moderately and positively correlated (r = .47, p < .01), indicating that these monitoring strategies tend to co-occur within family contexts.
A different pattern emerged for parent-reported data. Inconsistent parental monitoring was positively associated with externalising behaviours (r = .19, p < .01), whereas no significant associations were observed between restrictive or active monitoring and parent-reported behavioural outcomes. As expected, externalising and internalising behaviours were positively correlated within both child-reported (r = .61, p < .01) and parent-reported measures (r = .59, p < .01), reflecting the frequent co-occurrence of emotional and behavioural difficulties across informants.
Regression models for internalising and externalising behaviours
To examine the unique contribution of parental media monitoring strategies to children's behavioural adjustment, a series of multiple linear regression analyses were conducted. Demographic variables (sex, age, and socioeconomic indicators) and parental media monitoring strategies were entered simultaneously as explanatory variables in each model. This approach allowed us to examine the unique association of each parental monitoring strategy with children's behavioural adjustment while adjusting for demographic variables.
Tables 3 and 4 present the final models for externalising and internalising behaviours, integrating both child- and parent-reported data.
Multiple linear regression models for externalising behaviours
Multiple linear regression models for internalising behaviours
Regarding externalising behaviours, the child-reported model was statistically significant, explaining approximately 6% of the variance (R2 = .06, p < .001). Within this model, being male (β = −.11, p = .03), higher perceived family income (β = .10, p = .04), and inconsistent parental monitoring (β = .16, p = .00) were significantly associated with higher externalising behaviours. Similar findings emerged in the parent-reported model, which was also statistically significant and explained approximately 13% of the variance (R2 = .13, p < .001). In this model, being male (β = −.15, p = .00), higher perceived social class (β = .19, p = .00), and inconsistent parental monitoring (β = .16, p = .00) were significantly associated with externalising difficulties. Although the explained variance was modest, inconsistent parental monitoring remained significantly associated with externalising behaviours across both informants, suggesting a relatively stable association across reports.
Regarding internalising behaviours, the child-reported model was statistically significant, explaining approximately 7% of the variance (R2 = .07, p < .001). Within this model, being female (β = .14, p = .00), higher perceived family income (β = .13, p = .01), and inconsistent parental monitoring (β = .14, p = .01) were significantly associated with increased self-reported internalising symptoms. The parent-reported model was also statistically significant, explaining approximately 8.5% of the variance (R2 = .09, p < .001). In this model, perceived social class was the only significant variable explaining internalising behaviours (β = .18, p = .01). Inconsistent parental monitoring was not significantly associated with parent-reported internalising behaviours (β = .04, p = .47). Overall, inconsistent parental monitoring was associated with internalising behaviours only in the child-reported model, suggesting that children's perceptions of inconsistent supervision may be more closely related to their emotional difficulties than parents’ perceptions of these behaviours.
Cross-informant agreement on children's externalising and internalising behaviours
The results presented above pointed to differences between parent and child ratings, thus justifying the development of a cross-informant agreement analysis regarding children's internalising and externalising behaviours. Table 5 presents Pearson correlation coefficients (r), t values, and effect sizes used to examine levels of agreement and discrepancies between children's and parents’ reports of internalising and externalising behaviours. According to Cohen's (1988) guidelines for interpreting correlation magnitudes, the results indicated moderate agreement between informants for both externalising behaviours (r = .36, p < .01) and internalising behaviours (r = .35, p < .01).
Parent–child correlations, t-scores and effect sizes
Despite the positive and statistically significant correlations between informants, indicating moderate agreement, results also revealed significant differences between children's and parents’ ratings across both behavioural domains. Specifically, regarding externalising behaviours, statistically significant differences were observed between reports, t(13.51), p < .01, with a large effect size (d = 0.70), indicating substantial discrepancy between informants. For internalising behaviours, significant differences were also found, t(7.78), p < .01, although with a moderate effect size (d = 0.40). Across both domains, children reported higher levels of behavioural problems than did parents, as shown in Table 5.
Discussion
The present study examined the associations between parental media monitoring strategies and children's internalising and externalising behaviours within a community sample of Portuguese middle-school children. By shifting the analytical focus from mere digital access to the quality of parenting practices, the findings contribute to the growing body of literature suggesting that the consistency of parental involvement may be more relevant for children's behavioural adjustment than access to digital devices per se (Sanders et al., 2022).
The findings suggest partially distinct pathways through which parental monitoring strategies may relate to children's behavioural adjustment. Regarding externalising behaviours, inconsistent parental monitoring was associated with higher levels of behavioural difficulties across both child- and parent-reported models. From a developmental perspective, inconsistent monitoring may reduce the predictability of behavioural expectations and consequences, thereby limiting opportunities for children to develop stable self-regulatory patterns (Eisenberg et al., 2010; Sanders et al., 2022). In contrast, internalising difficulties may be more strongly associated with emotional insecurity and unpredictability within caregiving environments, particularly in digital contexts characterised by high accessibility and reduced external structure (Compas et al., 2017).
Across both child-reported data and parent-reported externalising outcomes, inconsistent parental monitoring remained significantly associated with behavioural difficulties after accounting for demographic variables (sex, age, and socioeconomic indicators). This finding is consistent with previous studies emphasising the importance of consistency and predictability in children's emotional and behavioural adjustment (Padilla-Walker et al., 2018). Importantly, inconsistent monitoring should not be interpreted merely as low parental involvement, but rather as an irregular pattern of rule enforcement in which parental responses fluctuate across situations. Such unpredictability may interfere with children's development of stable behavioural expectations regarding digital use and may be associated with difficulties in emotional and behavioural regulation (Coyne et al., 2021b).
These findings may also help explain why restrictive and active monitoring strategies showed weaker associations with behavioural outcomes in the multivariate models. In the current sample, the consistency and predictability of parental supervision appeared to be more strongly associated with behavioural adjustment than the specific content of parental mediation itself (Coyne et al., 2021a; Sanders et al., 2022). Although restrictive monitoring showed a weak negative bivariate association with internalising symptoms, this association did not remain significant once other variables were included in the regression models. This suggests that the presence of rules alone may be insufficient if those rules are not implemented consistently and predictably across contexts (Padilla-Walker et al., 2018; Pinquart, 2021).
Similarly, active parental monitoring was not significantly associated with either internalising or externalising behaviours. Although previous literature has frequently associated active mediation with positive developmental outcomes (e.g., Collier et al., 2016; Valkenburg et al., 2013), the present findings may reflect developmental changes occurring during late childhood and preadolescence. At this stage, parental discussions about media use may be perceived differently depending on their frequency, quality, or developmental appropriateness (Coyne et al., 2023a; Valkenburg et al., 2013). In addition, the benefits of active mediation may be more strongly associated with longer-term outcomes, such as critical thinking, emotional regulation, or media literacy, which may not be fully captured within cross-sectional behavioural assessments (Collier et al., 2016; Padilla-Walker et al., 2020).
Although the regression models explained a relatively modest proportion of variance (approximately 6%–13%), this pattern is consistent with previous research demonstrating that behavioural adjustment is a multidetermined phenomenon influenced by multiple individual, familial, and contextual factors (Valkenburg & Peter, 2013). Variables such as temperament, peer relationships, family emotional climate, and broader psychosocial stressors likely contribute substantially to children's behavioural functioning beyond the parental monitoring strategies assessed in the present study (Compas et al., 2017; Crone & Konijn, 2018; Valkenburg & Peter, 2013). Nevertheless, the fact that inconsistent parental monitoring remained significantly associated with behavioural difficulties across models and informants suggests that it may represent a relevant relational process within children's digital environments (Coyne et al., 2021a; Sanders et al., 2022).
Gender-related patterns observed in the present study followed well-established trends, with boys presenting higher levels of externalising behaviours and girls reporting higher levels of internalising symptoms. These differences are consistent with previous literature suggesting that emotional and behavioural difficulties may be expressed differently according to gender-related developmental and socialisation processes (Polanczyk et al., 2015).
The present study also highlighted meaningful discrepancies between parent and child reports. Although moderate levels of agreement emerged across informants, children consistently reported higher levels of behavioural difficulties than did parents, particularly regarding internalising symptoms. These discrepancies are consistent with multi-informant literature suggesting that different informants provide complementary rather than interchangeable perspectives on child functioning (Santos et al., 2020). Internalising difficulties may be less readily observable to caregivers due to their intrapersonal and covert nature, increasing the likelihood of under-identification when relying exclusively on parental reports. Differences between informants may additionally reflect variations in contextual observation, parental interpretation of behaviours, developmental expectations, and social desirability processes (Santos et al., 2020). Accordingly, the present findings reinforce the importance of integrating multiple sources of information when assessing behavioural adjustment during late childhood.
In conclusion, the present findings suggest that within contemporary digital environments, the consistency and predictability of parental media monitoring may be more strongly associated with children's behavioural adjustment than media restriction alone (Coyne et al., 2023b; Sanders et al., 2022). Although the observed associations were modest and causal conclusions cannot be drawn, inconsistent parental monitoring emerged as a relatively consistent correlate of behavioural difficulties across informants. These findings highlight the importance of considering family relational processes when examining children's adjustment in increasingly digitalised developmental contexts (Bronfenbrenner & Morris, 2006; Valkenburg & Piotrowski, 2017).
Limitations and directions for future research
Despite its contributions, the present study is subject to several limitations that must be acknowledged. First, the use of a convenience sample of 10- to 12-year-old children from a single geographic region limits the generalisability of the findings. Future research should employ probabilistic sampling designs across diverse sociocultural contexts to enhance external validity and ensure that the observed patterns are representative of broader populations (Bornstein et al., 2013). Furthermore, the relatively low levels of behavioural problems observed in this community sample may have resulted in a restriction of range, potentially attenuating the strength of the observed associations. While community and clinical samples are conceptually and methodologically distinct, future studies would benefit from oversampling populations at higher risk for behavioural maladjustment to determine the boundary conditions of the current findings.
Furthermore, the modest percentage of explained variance in the regression models indicates that the tested frameworks are not exhaustive. As suggested by the multidetermined nature of child development, factors such as individual temperament, the quality of the parent–child dyadic relationship outside the digital sphere, and school-level social influences likely exert a substantial influence on these behaviours (Coyne et al., 2020; Padilla-Walker et al., 2020). Additionally, the predominance of maternal reports in the current sample precluded a comparative examination of maternal versus paternal monitoring. Given that mothers and fathers often employ distinct mediation styles with varying developmental impacts, the inclusion of multiple caregivers is essential for future research to capture these nuances (Sanders et al., 2022). Furthermore, while child sex was included as a control variable in the current regression models, future research should move beyond this control-based approach. We recommend that subsequent studies conduct mean difference analyses and multi-group comparisons to explore sex-specific patterns more deeply. Investigating whether the predictive value of different parental monitoring strategies varies between boys and girls could provide more nuanced insights for tailored parenting interventions.
From a methodological standpoint, the reliance on self-report and proxy-report measures introduces the risk of social desirability bias and common-method variance (Podsakoff et al., 2003). Although the multi-informant design (integrating both parent and child reports) partially mitigates these concerns, future research would benefit from incorporating objective measures, such as observational data or digital trace data (e.g., screen time logs), to enhance measurement precision (Boers et al., 2020).
Finally, the cross-sectional design precludes conclusions regarding the directionality of the observed associations. Consistent with developmental-transactional perspectives, it remains possible that parental inconsistency and children's behavioural difficulties mutually influence one another over time (Pinquart, 2021). Future longitudinal research is needed to clarify these reciprocal dynamics and examine how parental monitoring and behavioural adjustment co-evolve across development.
Implications
Despite its limitations, the present study offers relevant theoretical and practical contributions to the fields of media psychology and developmental criminology. The findings suggest that within increasingly digitalised developmental environments, the consistency and predictability of parental media monitoring may be more strongly associated with children's behavioural adjustment than media restriction alone (Livingstone & Blum-Ross, 2020; Sanders et al., 2022). In this sense, the present study contributes to a growing body of literature emphasising the importance of relational and contextual parenting processes in understanding children's digital experiences.
From a practical standpoint, the findings suggest the importance of moving beyond exclusively restrictive or gatekeeping approaches to digital parenting toward supervision strategies characterised by clearer, more consistent, and developmentally appropriate behavioural expectations. Rather than focusing solely on limiting screen time or prohibiting specific forms of digital engagement, effective parental mediation may depend more strongly on the establishment of predictable and coherent monitoring practices across situations (Padilla-Walker et al., 2018). Such consistency may help provide children with a more stable regulatory environment associated with behavioural and emotional adjustment (Eisenberg et al., 2010).
These findings may hold relevance within the Portuguese context, where evidence-based guidelines regarding digital parenting practices are still relatively limited (Canário et al., 2022). Parental education and prevention programmes may therefore benefit from incorporating content focused not only on digital risks, but also on the psychological importance of consistency, communication quality, and predictable rule enforcement within family environments. Supporting caregivers in maintaining coherent supervision practices under conditions of stress and increasing digital accessibility may represent a potentially important target for intervention efforts.
From an educational perspective, the present findings also reinforce the value of integrating media literacy and family-based digital education initiatives within school settings. Strengthening collaboration between schools and families may contribute to more coherent and developmentally appropriate approaches to children's media use across contexts.
Finally, the multi-informant design adopted in the present study highlights the importance of integrating both child and caregiver perspectives when assessing behavioural adjustment in digital contexts. The meaningful discrepancies observed between informants reinforce previous literature suggesting that different sources of information provide complementary rather than interchangeable insights into children's emotional and behavioural functioning (Achenbach, 2006; De Los Reyes et al., 2015). Collectively, these findings emphasise the importance of examining children's digital experiences within broader family relational contexts rather than treating media exposure as an isolated behavioural risk factor.
Footnotes
Ethical Consideration
The authors declare that the study was approved by the Ethics Committee of the Faculty of Law, University of Porto, Portugal (approval no. 11/2023 of 18th December 2023). Furthermore, the study received approval from the Directorate General for Education through its monitorisation platform. Informed consent was obtained from all participants, and the necessary authorisations were obtained before data collection procedures.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability statement
Not applicable.