Emotional and Behavioral Difficulties and Academic Motivation in Rural Adolescents: An Ecological-Psychosocial Model

1.1. Theoretical Framework and Conceptual Model

This study was guided by an ecological-psychosocial framework that integrates three complementary theoretical perspectives. First, ecological systems theory (Bronfenbrenner) [16] emphasizes that adolescent development is shaped by multiple interrelated contexts, including individual characteristics, family processes, peer relationships, school experiences, and broader sociocultural conditions. From this perspective, screen time behaviors, parenting styles, caregiver distress, and social connectedness can be understood as ecological factors that may influence adolescent psychosocial and academic functioning.

Second, the dual-factor model of mental health (Keyes; Suldo & Shaffer) [8, 9] provides a framework for understanding adolescent adjustment through both negative and positive indicators of functioning. This model distinguishes psychological distress from well-being and suggests that the absence of symptoms does not necessarily indicate optimal mental health. In the present study, emotional and behavioral difficulties, youth distress, well-being, and prosocial behavior were therefore included to capture both risk and strength dimensions of adolescent psychosocial functioning.

Third, self-determination theory (Deci & Ryan) [10, 11] was used to frame academic motivation as an outcome shaped by adolescents' psychological functioning and social context. According to this theory, motivation is supported when adolescents experience competence, autonomy, and relatedness. Emotional and behavioral difficulties may undermine academic motivation by reducing students' capacity for self-regulation, perceived competence, and positive engagement in school-related activities. Conversely, prosocial behavior and social connectedness may support academic motivation by strengthening relatedness and adaptive social functioning.

Based on these perspectives, the present study conceptualized screen time behaviors as behavioral exposures, emotional and behavioral difficulties as a potential psychosocial mechanism, and academic motivation as the academic outcome. Additional ecological -psychosocial correlates, including parenting styles, caregiver distress, youth distress, well-being, social connectedness, and prosocial behavior, were included to provide a more integrated account of adolescent functioning. The proposed model assumes that screen time behaviors may be associated with academic motivation both directly and indirectly through emotional and behavioral difficulties, while family and psychosocial factors may help explain variations in adolescent adjustment and motivation. Figure 1 presents the conceptual framework guiding the analysis.

1.2. The Present Study and Hypotheses

The present study addresses this gap by examining the associations between parenting styles, screen time behaviors, psychosocial functioning, and academic motivation among adolescents within a dyadic framework that incorporates both adolescent and caregiver variables. Given the complexity of these relationships and the characteristics of the available sample, the analysis was conducted with an exploratory orientation, aiming to identify patterns of association rather than test a fully specified causal model.

Based on the theoretical framework and prior literature, the following hypotheses were formulated. Given the cross-sectional design and modest sample size, these hypotheses were tested within an exploratory analytical framework.

H1: Higher levels of screen time, particularly recreational screen use, were expected to be associated with higher emotional and behavioral difficulties.

H2: Higher emotional and behavioral difficulties were expected to be associated with lower academic motivation.

H3: Emotional and behavioral difficulties were expected to mediate the association between screen time behaviors and academic motivation.

H4: Higher youth psychological distress and caregiver distress were expected to be associated with greater emotional and behavioral difficulties.

H5: Higher well-being and authoritative parenting were expected to be associated with lower emotional and behavioral difficulties, whereas higher prosocial behavior and social connectedness were expected to be associated with higher academic motivation.

2.1. Study Design and Setting

This study employed a cross-sectional design to examine associations between parenting styles, screen time behaviors, psychosocial functioning, and academic motivation among adolescents. Data were collected at a public junior high school in Salatiga, Central Java, Indonesia, representing a rural educational context. The analytical approach was framed as exploratory given the sample size relative to model complexity.

2.2. Participants and Sampling

The study included 92 adolescent-caregiver dyads. Each adolescent was paired with one primary caregiver, defined as an adult responsible for daily care and supervision and who had lived with the adolescent for at least six months.

Participants were recruited from students in grades 7 to 9 using an alternating sampling approach. Adolescents completed questionnaires in a supervised classroom setting, while caregivers completed their questionnaires independently after obtaining informed consent.

Adolescents were eligible if they were enrolled at the selected school, aged 12 to 16 years, living with a primary caregiver, and able to complete the questionnaire independently. Exclusion criteria included severe cognitive or developmental conditions that could interfere with participation or absence during data collection. Caregivers were eligible if they met the definition of a primary caregiver and agreed to participate.

A dyadic design enabled the inclusion of caregiver-level variables alongside adolescent outcomes within a single analytic framework.

2.3. Sample Size Justification

Sample size adequacy was assessed according to the 10-times rule for PLS-SEM, which recommends a minimum sample size equal to 10-times the maximum number of predictor constructs directed toward an endogenous variable. The proposed model included 8 latent constructs (6 adolescent measures and 2 caregiver measures), resulting in a minimum sample size requirement of 80 participants (10 × 8 = 80). The study recruited 92 adolescent-caregiver dyads, thereby exceeding the recommended threshold. In addition, statistical inference was strengthened through a non-parametric bootstrapping procedure with 5,000 resamples, which was used to estimate standard errors, confidence intervals, and the significance of the structural path coefficients.

2.4. Research Procedures

Prior to data collection, permission was obtained from the school administration. Study objectives and procedures were explained to both adolescents and caregivers. Written informed consent was obtained from caregivers, and assent was obtained from adolescents.

Data were collected on-site. Adolescents completed questionnaires in a supervised classroom setting, while caregivers completed their questionnaires independently at the same location. No identifying information was collected, and responses were recorded anonymously. All data were handled confidentially in accordance with ethical standards for research involving human participants.

2.5. Measures

2.6. Psychometric Evaluation and Instrument Transparency

To ensure the instrument's contextual validity, psychometric evaluations were performed specifically for the current rural sample. Internal consistency was evaluated using Cronbach's alpha based on the present sample, with reliability estimates interpreted using conventional thresholds (α ≥ 0.70). Item-total correlations were examined to assess item-level performance.

Detailed information on the questionnaire package is provided in Appendix Table A2, including respondent type, construct coverage, number of items, response format, scoring procedure, and representative items for each measure. Where full item reproduction was permitted, item lists are provided (e.g., Appendix Table A4 for SCS-R). For standardized instruments with copyright or use restrictions, representative items and scoring rules are reported to support transparency and reproducibility. A new Appendix Table A5 presents item-to-construct mapping for all variables included in the structural model.

Reliability was evaluated empirically within the current dataset rather than assumed from prior studies. As shown in Appendix Table A3, several items, particularly within the social connectedness scale, showed weak or negative item-total correlations. Results involving this construct were interpreted cautiously.

2.7. Data Analysis

All analyses were conducted using Python with the pandas, SciPy, and StatsModels libraries. Descriptive statistics were computed for all variables. Due to non-normality in several measures, associations were examined using Spearman's rank correlation. Multicollinearity was assessed using variance inflation factors (VIF), with all values below conventional thresholds.

Regression-based path modeling was conducted using ordinary least squares estimation. Altogether, two models were specified: one predicting SDQ difficulties and one predicting academic motivation. Predictor selection was based on theoretical relevance and model parsimony.

Bootstrap resampling (5,000 iterations) was used to estimate coefficients and 95% confidence intervals. Indirect effects were examined using bootstrap-based mediation analysis. Effects were considered supported when confidence intervals did not include zero. Mediation results were interpreted cautiously given the exploratory design.

2.8. Data Handling and Statistical Thresholds

Analyses were conducted using complete-case data, with missing values handled through listwise deletion. Missing data were minimal and did not materially affect the sample. Statistical significance was defined as p < .05. Interpretation emphasized effect sizes and confidence intervals rather than reliance on p-values alone.

3.1. Sociodemographic Characteristics

The sample consisted of 92 adolescent-caregiver dyads. Adolescents were nearly evenly distributed by gender, and caregivers were predominantly female. Most caregivers were married. Detailed characteristics are reported in Table 1.

3.2. Descriptive Statistics

Descriptive statistics for all study variables are presented in Table 2. Academic motivation, social connectedness, and SDQ difficulties showed moderate dispersion. Screen time varied across activity types, with academic-related use higher than other categories.

3.3. Correlation Analysis

Spearman correlation results are presented in Table 4. SDQ difficulties were strongly associated with youth distress and inversely related to well-being and social connectedness. Academic motivation was positively associated with prosocial behavior and social connectedness, and inversely related to SDQ difficulties.

Screen time variables showed smaller and less consistent associations compared to psychosocial variables.

3.4. Direct Effects

All variance inflation factors were below conventional thresholds, indicating no evidence of problematic multicollinearity (see Appendix Table A1). Bootstrapped regression results are presented in Table 5.

3.5. Indirect Effects

Bootstrapped indirect effects are presented in Table 6. Most indirect effects through SDQ difficulties were not statistically supported. A positive indirect effect was observed for gaming via SDQ difficulties, although the direction and magnitude warrant cautious interpretation.

Overall, mediation patterns were limited and exploratory.

4.1. Overview of Findings

The present study examined associations between parenting styles, screen time behaviors, psychosocial functioning, and academic motivation within an exploratory framework. Overall, the findings indicate that emotional and behavioral difficulties were more strongly associated with psychological factors, particularly distress and well-being, than with behavioral variables such as screen time. Academic motivation was negatively associated with emotional and behavioral difficulties and positively associated with prosocial behavior, while screen-based activities showed relatively weak and inconsistent relationships with the outcomes. Mediation effects were limited and did not provide consistent support for indirect pathways.

4.2. Interpretation of Key Associations

The strong association between psychological distress and emotional and behavioral difficulties is consistent with the broader literature on adolescent mental health, which conceptualizes distress as a central correlate of internalizing and externalizing problems during a period of heightened developmental sensitivity [28]. At the same time, the inverse association between well-being and difficulties aligns with the dual-factor model of mental health, which emphasizes that positive functioning and psychopathology represent related but distinct dimensions [8, 9]. The present findings support this perspective, suggesting that both distress and well-being contribute uniquely to adolescents' psychosocial functioning.

Academic motivation showed a pattern consistent with motivational theories, particularly self-determination theory. The negative association between emotional and behavioral difficulties and motivation suggests that dysregulation and behavioral problems may interfere with students' engagement, perceived competence, and persistence in school-related activities [10, 11]. In contrast, the positive association between prosocial behavior and motivation may reflect the role of social competence and relatedness in supporting academic engagement [10, 11]. Although Anvari's recent review focuses primarily on motivation in performance-related organizational contexts, its emphasis on motivation as a driver of effort, persistence, and goal-directed behavior is consistent with the present interpretation of academic motivation as a performance-relevant educational construct [29]. These findings indicate that motivation is not only an individual cognitive construct but is also linked to broader psychosocial functioning.

In contrast, screen time variables showed relatively weak and inconsistent associations with emotional and behavioral difficulties and academic motivation. This finding does not necessarily indicate that screen use is unrelated to adolescent functioning. Rather, it suggests that broad screen-time indicators may not fully capture the conditions under which digital media use becomes beneficial, neutral, or harmful. Prior research has indicated that the effects of screen-time may depend on content, purpose, timing, social context, and whether screen use displaces sleep, physical activity, face-to-face interaction, or academic engagement [12, 13]. Since the present study measured screen time by general activity domains without assessing more specific features such as nighttime use, passive versus active engagement, emotional tone of content, parental mediation, or problematic use patterns, the findings should be interpreted cautiously.

4.3. Limited Mediation Effects and Model Complexity

Contrary to expectations, the mediation analyses did not provide consistent support for indirect pathways linking screen time or parenting variables to outcomes through emotional and behavioral difficulties. Most indirect effects were small and not statistically supported, and thus should be interpreted with appropriate caution.

Several factors may explain these findings. First, the relatively modest sample size limits statistical power for detecting indirect effects, which are typically smaller and more sensitive to sampling variability [30, 31]. Second, the complexity of the initial conceptual model relative to the available data may have reduced the stability of parameter estimates. Third, measurement limitations, particularly variability in item performance within some scales, may have attenuated the detection of indirect effects.

The findings do not support strong mediation mechanisms in the present dataset. Instead, they suggest that the observed relationships are better interpreted as direct associations within an exploratory framework rather than as evidence of well-defined causal pathways.

4.4. Measurement Considerations

The psychometric evaluation indicated that most instruments demonstrated acceptable internal consistency within the present sample. However, the social connectedness scale showed relatively low reliability, accompanied by substantial variability in item-total correlations, including several weak or negative values.

This pattern may reflect challenges in construct alignment within the current context, including potential issues related to cultural adaptation, item interpretation, or multidimensionality. Previous research has emphasized that the psychometric performance of established instruments may vary across populations and settings, particularly when applied in different cultural contexts [32, 33]. As a result, findings involving social connectedness should be interpreted cautiously, and future studies should consider further validation or refinement of this measure.

4.5. Theoretical Implications

The findings offer several theoretical implications. First, they provide partial support for an ecological-psychosocial understanding of adolescent academic motivation by showing that motivational functioning is more closely linked to emotional and behavioral adjustment than to screen time alone. This suggests that screen exposure should not be treated as an isolated behavioral risk factor but should be interpreted within broader psychological and relational contexts.

Second, the findings support the dual-factor model of mental health by showing that psychological distress and well-being were both associated with emotional and behavioral difficulties. This pattern indicates that adolescent adjustment cannot be fully understood by measuring symptoms alone. Positive well-being may provide additional explanatory value in understanding why some adolescents show fewer difficulties despite exposure to contextual or behavioral risks.

Third, the findings refine the proposed mediation model. Although emotional and behavioral difficulties were associated with academic motivation, they did not consistently mediate the relationship between screen time and motivation. This suggests that the screen time-motivation link may be more indirect, context-dependent, or contingent on unmeasured factors such as sleep quality, digital content, parental mediation, self-regulation, and perceived academic competence.

4.6. Practical Implications

The findings have practical implications for school-based mental health and academic support. First, schools should not rely solely on screen time reduction as a strategy to improve academic motivation. Although excessive screen use may be relevant, the present findings suggest that emotional distress, well-being, prosocial behavior, and behavioral adjustment may be more immediate targets for intervention.

Second, school-based screening should include both negative and positive indicators of mental health. Assessing distress alone may overlook students with low well-being, poor social connectedness, or reduced motivation. A brief screening model combining emotional difficulties, well-being, and academic motivation may assist teachers and school counselors in identifying students who require early support.

Third, interventions to strengthen academic motivation should incorporate psychosocial components. Programs that promote emotion regulation, peer support, prosocial behavior, and supportive teacher-student relationships may be more effective than interventions focused only on study habits or digital restriction. For rural adolescents, school-based approaches may be particularly important, as access to specialized psychological services can be limited.

Fourth, parents and caregivers should be involved in prevention efforts. Rather than framing screen use only as a disciplinary issue, caregivers can be guided to support balanced digital habits, monitor emotional changes, encourage sleep routines, and maintain warm communication with adolescents.

4.7. Limitations and Future Directions

Several limitations should be noted. The cross-sectional design does not allow conclusions about directionality, and the observed associations should be interpreted accordingly. Estimates of indirect effects derived from cross-sectional data may also be biased when interpreted as longitudinal processes [34]. Longitudinal designs would be needed to examine how these relationships unfold over time.

The sample size was below the recommended threshold for the model's complexity, potentially reducing statistical power, particularly for detecting indirect effects. Estimates from the present analyses should therefore be interpreted with caution.

Measurement issues also need to be considered. In particular, the relatively low reliability of the social connectedness scale in this sample may have attenuated some associations. Further evaluation of this measure in similar populations is needed.

The study was conducted in a single school context, which may limit the generalizability of the findings. Replication in more diverse educational and cultural settings would help clarify the extent to which these patterns hold across contexts.

Future work would benefit from longitudinal designs, improved measurement strategies, and more parsimonious models to better capture the relationships among parenting, screen use, psychosocial functioning, and academic outcomes.