青春期是情绪和行为问题的关键发展阶段，许多心理健康问题在这一时期首次出现或显著恶化。其中，内化问题和外化问题是青少年群体中普遍存在的心理病理症状。青少年内外化问题机制模型指出情绪是内外化问题的直接风险因素。情绪反应性全面地表征了情绪在各方面的变化。因此亟需研究在理论指导下，从实证角度检验青少年情绪反应性与内外化问题的关系。基于动态系统理论方法论的指导，对青少年情绪反应性的研究需要关注到其在微观和宏观不同层面上的变化。因此，青少年情绪反应性和内外化问题之间的关系同样需要在微观和宏观的不同层面上被检验。此外，青少年内外化问题机制模型表明同伴侵害是青少年发展的重要环境因素，会影响情绪反应性和内外化问题之间的关系。结合上述理论对研究问题的指导，本研究共开展了三项实证研究（5个子研究），具体如下： 研究一在静态层面中，从宏观横断特征的视角对青少年进行横断的大样本问卷调查（N = 2108），初步探索了青少年情绪反应性与内化问题和外化问题的关系。结果显示，青少年情绪反应性及其各维度与内化问题和外化问题存在显著的正相关（ps < .01）。回归分析结果进一步显示，情绪敏感、情绪强度和情绪惰性能够正向预测内化问题；情绪敏感、情绪强度能够正向预测外化问题。 研究二在动态层面中，从微观瞬时变化的视角检验青少年情绪反应性、内化问题、外化问题的短期关系，包含两个子研究。研究2a对180名青少年进行了为期七天、每天三次的动态评估研究，建立多层线性模型以检验日常生活情境下，青少年情绪反应性、内化问题、外化问题的瞬时中介模型。中介分析结果显示，情绪敏感/情绪强度能够通过正向预测内化问题进而正向预测外化问题（Est. = 0.06 / 0.11, p < .001 / <.001, 95% CI = [0.034, 0.080] / [0.091, 0.138])。而情绪惰性、内化问题、外化问题的瞬时中介模型不成立。研究2b对193名青少年进行了同样设计的动态评估研究和数据分析，检验日常生活情境中，同伴侵害在情绪反应性、内化问题、外化问题的瞬时中介模型中的调节作用。中介分析结果显示，情绪敏感/情绪强度能够通过正向预测内化问题进而正向预测外化问题（Est. = 0.02 / 0.04, p = .02 / < .001, 95% CI = [0.003, 0.033] / [0.019, 0.050]）；情绪惰性、内化问题、外化问题的瞬时中介模型不成立。调节分析结果显示，同伴侵害在情绪敏感/情绪强度、内化问题、外化问题的瞬时中介模型中具有显著的调节作用。简单斜率分析表明，在情绪敏感→内化问题→外化问题的瞬时中介模型中，间接效应在同伴侵害水平较低（-1SD; Est. = 0.02, p = .03, 95% CI = [0.001, 0.029]）、适中（Mean; Est. = 0.02, p = .02, 95% CI = [0.001, 0.030]）和较高（+1SD; Est. = 0.02, p = .02, 95% CI = [0.001, 0.032]）时逐渐增强。在情绪强度→内化问题→外化问题的瞬时中介模型中，间接效应在同伴侵害水平较低（-1SD; Est. = 0.03, p < .001, 95% CI = [0.016, 0.044]）、适中（Mean; Est. = 0.03, p < .001, 95% CI = [0.017, 0.046]）和较高（+1SD; Est. = 0.03, p < .001, 95% CI = [0.018, 0.048]）时逐渐增强。 研究三在动态层面中，从宏观发展变化的视角检验青少年情绪反应性、内化问题、外化问题的长期关系，包含两个子研究。研究3a对186名青少年进行了三年的追踪研究，以检验青少年情绪反应性、内化问题、外化问题的纵向中介模型，以及同伴侵害在模型中的调节作用。结果发现T1情绪敏感/情绪强度/情绪惰性、T2内化问题、T3外化问题的纵向中介模型均不成立。研究3b对351名青少年进行了三次追踪研究，建立潜变量增长模型、中介模型和调节模型，以检验青少年情绪反应性、内化问题、外化问题的发展轨迹中介模型。中介分析结果显示，情绪敏感/情绪强度/情绪惰性截距可以通过显著正向预测内化问题截距，进而预测外化问题截距（间接效应 = 2.03 / 2.58 / 1.90, p = .03 / .01 / .03, 95% CI = [1.200, 5.398] / [1.477, 5.841] / [1.156, 5.272]）。情绪敏感/情绪强度斜率可以通过显著正向预测内化问题斜率，进而预测外化问题斜率（间接效应 = 2.98 / 3.89, p = .04 / .02, 95% CI = [1.576, 8.031] / [2.055, 10.521]）。调节分析结果显示，同伴侵害的发展在发展轨迹中介模型中存在调节作用。简单斜率分析表明，在情绪敏感斜率→内化问题斜率→外化问题斜率的发展轨迹中介模型中，间接效应在同伴侵害斜率较低（-1SD; Est. = 1.20, p < .001, 95% CI = [0.614, 1.802]）、适中（Mean; Est. = 0.50, p < .05, 95% CI = [0.091, 0.863]）和较高（+1SD; Est. = -0.20, p > .05, 95% CI = [-0.812, 0.291]）时逐渐减弱。 整体而言，本研究基于动态系统理论方法论的指导，遵循“发现现象”到“解释现象”的思路，从静态层面和动态层面（微观瞬时变化和宏观发展变化）不同角度，运用多种设计逐渐深入地检验了青少年情绪反应性、内化问题、外化问题之间的关系，以及同伴侵害的调节作用。本研究从多个层面、视角揭示了青少年内化问题和外化问题的发展机制和形成路径，验证和深化了青少年内外化问题机制模型的理论框架。研究结果为素质—压力模型、行为不当模型和差别易感性模型在微观层面上的运用提供了支持，有助于推进相关领域的理论发展，对临床干预有重要的启示和价值。

Adolescence is a critical developmental stage for emotional and behavioral problems, during which many mental health issues first emerge or significantly worsen. Among these, internalizing and externalizing problems are common psychological pathologies in adolescents. The model for adolescent internalizing and externalizing problems suggests that emotion is a direct risk factor for these problems. The concept of emotion reactivity comprehensively characterizes changes in emotions. Therefore, it’s essential to empirically examine the relationship between adolescent emotion reactivity and internalizing and externalizing problems while adhering to theoretical guidance. Guided by the methodology of Dynamic System Theory (DST), research on adolescent emotional reactivity should attentively consider its changes at both micro and macro levels. Consequently, the relationship between adolescent emotional reactivity and internalizing and externalizing problems also needs to be examined at various levels. Additionally, the model for adolescent internalizing and externalizing problems indicates that peer victimization is an important environmental factor in adolescent development, affecting the relationship between emotional reactivity and internalizing and externalizing problems. Combining the above research questions, this study has conducted three empirical studies (including five sub-studies), specifically as follows: In Study 1, from a static perspective, we conducted a cross-sectional survey on a large sample of adolescents (N = 2108) to preliminarily explore the relationship between adolescents’ emotion reactivity and internalizing and externalizing problems. The results showed that there is a significant positive correlation (p < .01) between adolescents’ emotion reactivity and its dimensions with internalizing and externalizing problems. Regression analysis revealed that factors such as emotion instability, emotion intensity, and emotion inertia were positive predictors of internalizing problems. Similarly, emotion instability and emotion intensity were positive predictors of externalizing problems.. In Study 2, from a dynamic perspective, we examined the short-term relationships among adolescents’ emotion reactivity, internalizing problems, and externalizing problems from the level of micro- momentary changes, comprising two sub-studies. In Study 2a, we conducted an Ambulatory Assessment (AA) study on 180 adolescents for seven days, three times a day, establishing a Multilevel Linear Model (MLM) to test the momentary mediation model of adolescents’ emotion reactivity, internalizing problems, and externalizing problems in daily life situations. Mediation analysis results showed that emotion instability/emotion intensity could positively predict internalizing problems and then positively predicted externalizing problems (Est. = 0.06 / 0.11, p < .001 / <.001, 95% CI = [0.034, 0.080] / [0.091, 0.138]). However, the momentary mediation model involving emotional inertia, internalizing problems, and externalizing problems was not supported. Study 2b employed the same design as the AA study and analysis on 193 adolescents, to test the moderating role of peer victimization in the momentary mediation model of emotion reactivity, internalizing problems, and externalizing problems in daily life situations. The mediation analysis results showed that internalizing problems mediated the relationship between emotion instability/emotion intensity and externalizing problems (Est. = 0.02 / 0.04, p = .02 / < .001, 95% CI = [0.003, 0.033] / [0.019, 0.050]); the momentary mediation model involving emotional inertia, internalizing problems, and externalizing problems was not supported. Moderation analysis results indicated that peer victimization had a significant moderating effect in the momentary mediation model of emotion instability/emotion intensity, internalizing problems, and externalizing problems. The simple slope analysis indicated that in the momentary mediation model of emotion instability→internalizing problems→externalizing problems, the indirect effect gradually strengthened when the level of peer victimization at low (-1SD; Est. = 0.02, p = .03, 95% CI = [0.001, 0.029]), moderate (Mean; Est. = 0.02, p = .02, 95% CI = [0.001, 0.030]), and high (+1SD; Est. = 0.02, p = .02, 95% CI = [0.001, 0.032]). In the momentary mediation model of emotion intensity→internalizing problems→externalizing problems, the indirect effect gradually strengthened when the level of peer victimization at low (-1SD; Est. = 0.03, p < .001, 95% CI = [0.016, 0.044]), moderate (Mean; Est. = 0.03, p < .001, 95% CI = [0.017, 0.046]), and high (+1SD; Est. = 0.03, p < .001, 95% CI = [0.018, 0.048]). In Study 3, from a dynamic perspective, we examined the long-term relationships among adolescents’ emotion reactivity, internalizing problems, and externalizing problems from the level of macro-developmental changes, comprising two sub-studies. Study 3a conducted a three-year longitudinal study on 186 adolescents to test the longitudinal mediation model of adolescents’ emotion reactivity, internalizing problems, and externalizing problems, as well as the moderating role of peer victimization in the model. The results found that the longitudinal mediation model involving T1 emotion instability/emotion intensity/emotion inertia, T2 internalizing problems, and T3 externalizing problems was not supported. Study 3b conducted a three-wave longitudinal study on 351 adolescents, establishing Latent Growth Model (LGM), mediation models, and moderation models to test the developmental trajectory mediation model of emotion reactivity, internalizing problems, and externalizing problems. Mediation analysis results showed that the intercepts of emotion instability/emotion intensity/emotion inertia could significantly positively predict the intercepts of internalizing problems, and then predicted the intercepts of externalizing problems (Indirect Effect = 2.03 / 2.58 / 1.90, p = .03 / .01 / .03, 95% CI = [1.200, 5.398] / [1.477, 5.841] / [1.156, 5.272]). The slopes of emotion instability/emotion intensity could significantly positively predict the slopes of internalizing problems, and then predicted the slopes of externalizing problems (Indirect Effect = 2.98 / 3.89, p = .04 / .02, 95% CI = [1.576, 8.031] / [2.055, 10.521]). Moderation analysis results indicated that the development of peer victimization had a moderating effect in the developmental trajectory mediation model. Simple slope analysis revealed that the predictive effect of the slope of emotion instability on the slope of externalizing problems gradually weakened at low (-1SD; Est. = 1.20, p < .001, 95% CI = [0.614, 1.802]), moderate (Mean; Est. = 0.50, p < .05, 95% CI = [0.091, 0.863]), and high (+1SD; Est. = -0.20, p > .05, 95% CI = [-0.812, 0.291]) levels of the slope of peer victimization. In summary, this research, grounded in the methodology of DST, adopts an approach that transitions from the discovery of phenomena to the explanation of phenomena. It employs diverse designs to examine the relationships among adolescents’ emotion reactivity, internalizing problems, and externalizing problems, as well as the moderating role of peer victimization, from various perspectives including static and dynamic dimensions (micro-momentary changes and macro-developmental changes). Additionally, it also investigates the interaction between peer victimization and relationships in these models. This comprehensive research across multiple dimensions and perspectives contributes to unraveling the developmental mechanisms and pathways of adolescents’ internalizing and externalizing problems. It advances the theoretical framework of the model for adolescent internalizing and externalizing problems. This study provides support for the application of the Diathesis-stress Interaction Model, the Acting Out Model, and the Differential Susceptibility Model at the micro-level. It contributes to the theoretical advancement and has significant implications for clinical interventions.