受欺凌者作为欺凌事件的主要受害群体，是欺凌研究中高度关注的群体。受欺凌者消极后果的有效干预受到群体内部异质性的影响，因此要对受欺凌亚型进行精准识别。根据社会角色理论，受欺凌角色的识别同时依赖于个体层面的主观认知和群体层面的客观认知，两者共同构建了受欺凌的全貌。但是，个体对于受欺凌的主观认知与群体的客观认知存在差异，由此存在不同的受欺凌亚型，如自我确认组（自我识别为受欺凌，同伴识别为未受欺凌）和同伴确认组（同伴识别为受欺凌，自我识别为未受欺凌）。有必要进一步明确不同受欺凌亚型的稳定性和发展变化及其影响因素，以便可以用于针对不同受欺凌群体实施相应的干预措施。此外，欺凌依赖于同伴群体关系，欺凌亚型的发展变化依赖于班级人际关系的发展变动。因此，本研究基于主客观视角进行青少年受欺凌亚型的识别，采用横断和纵向研究设计，以及跨群体比较，探究受欺凌亚型及其表现、发展模式和预测因素。

研究一采用横断设计，从主客观视角出发，通过中国青少年本土样本来考察是否存在不同受欺凌亚型。在此基础上，研究二采用纵向设计，对被试进行追踪测量，进而探究青少年早期受欺凌亚型的发展模式的稳定性和转变性，及内化问题和同伴地位对于转变的预测作用。基于研究二，研究三继续进行两次追踪调查，根据青少年是否换班划分为恒定群体和重建群体，比较和探究恒定和重建群体中受欺凌亚型及其发展模式和预测因素的差异。

被试来自中国中部某省会城市两个区县7所普通初中的学生。这7所学校包括2所城市学校、1所县城学校和4所乡镇学校。分别在七年级第一学期期末 (T1)、七年级第二学期期末 (T2)、八年级第一学期期末 (T3) 和八年级第二学期期末 (T4) 对这7所学校的47个班级进行了问卷调查。最终共收集到2138份有效问卷，被试首次参与调查 (T1) 的平均年龄为12.99岁，标准差为0.62，其中49.5%为男生。受欺凌分别使用自我报告法和同伴提名法获得，抑郁、自尊根据自我报告法获得，受欢迎度、朋友数量和同伴拒绝采用同伴提名方式获得。数据分析采用 SPSS 21.0和Mplus8.3。

研究的主要结果如下：

（1）基于主客观认知视角，青少年早期主要存在三种受欺凌亚型，分别是自我确认组(T1: 8.3%)、同伴确认组 (T1: 7.2%) 和未受欺凌组 (T1: 84.5%)。自我确认组的自我报告受欺凌水平较高、同伴报告受欺凌水平较低，而同伴确认组的同伴报告受欺凌水平较高，自我报告受欺凌水平较低。未受欺凌组的比例高于自我确认组和同伴确认组，表明大多数青少年较少受到欺凌。

（2）不同受欺凌亚型的发展模式存在差异。未受欺凌组的稳定性最高 (T1-T2: 92.8%)，而自我确认组 (T1-T2: 64.8%) 和同伴确认组 (T1-T2: 75.5%) 的稳定性相对较低。自我确认组和同伴确认组倾向于向未受欺凌组转变，说明着受欺凌情况有逐渐变好的趋势。

（3）内化问题和同伴地位对于不同亚型的转变具有预测作用。具体来说，内化问题（抑郁和低自尊）主要和自我确认组与其他组的转变相关联，表现为抑郁促使未受欺凌组向自我确认组转变，低自尊阻碍自我确认组向未受欺凌组转变。同伴拒绝主要和同伴确认组与其他组转变相关联，表现为同伴拒绝促进未受欺凌组和自我确认组向同伴确认组转变的概率，阻碍同伴确认组向其他组转变。

（4）在恒定群体 (T3-T4) 中，受欺凌亚型的稳定性和转变性有所变化，具体表现为自我确认组 (64.8% → 72.7%) 和同伴确认组 (75.5% → 96.6%) 的稳定性有所上升。其次，仅发现同伴拒绝促进未受欺凌组向同伴确认组转变，内化问题对于不同亚型的预测作用不显著。

（5）在重建群体 (T3-T4) 中，受欺凌亚型的稳定性与群体建立初期 (T1-T2) 基本一致 (同伴确认组: 75.5% → 73.5%; 未受欺凌组: 75.5% → 73.5%)。其次，内化问题和同伴地位对于不同亚型的预测作用仍然显著，同伴拒绝促进未受欺凌组向同伴确认组转变，此外，还发现同伴拒绝、抑郁促进未受欺凌组向自我确认组转变。

总的来说，本研究发现了三组受欺凌亚型，包括自我确认组、同伴确认组和未受欺凌组。其次，不同受欺凌亚型具有跨时间的稳定性，但仍有部分个体在一学期后转变为其他亚型群体，抑郁、自尊、同伴拒绝在转变过程中发挥作用。此外，恒定群体中受欺凌亚型的稳定性有所提高，但不同因素对于亚型转变的预测作用有所减弱，而重建群体中亚型的稳定性基本保持不变，且不同因素的预测作用依然显著。

Victims are a highly concerned group in bullying research due to their serious internalizing and externalizing problems. There is heterogeneity within the victims. To achieve effective intervention for the negative consequences of victimization, it is essential to accurately identify victim subtypes. According to the social role theory, the identification of victims depends on both self-perceptions at the individual level and peer-perceptions at the group level, which together construct the complete picture of victimization. However, there are differences between the self- and peer-perceptions of victimization. Based on these differences, different types of victims are identified such as self-identified victims (identified as victim by themselves but not by their peers) and peer-identified victims (identified as victim by their peers but not by themselves). It is essential to further clarify the stability and changes of different victim subtypes, as well as the factors that influence their transition in order to implement appropriate intervention measures targeted at different victim subtypes. Therefore, this study focuses on the identification of victim subtypes among teenagers, using a cross-sectional and longitudinal design to explore the victim subtypes and their victimization forms, developmental patterns, and predictive factors based on both self- and peer- perceptions.

In Study 1, a cross-sectional design was adopted to explore the victim subtypes and their victimization forms from both self- and peer-perceptions among teenagers. This study hopes to investigate whether different victim subtypes exist based on Chinese samples. Based on this, Study 2 is a longitudinal design that tracks the participants and explores the developmental patterns of victim subtypes, as well as the predictive effects of internalizing problems and peer status, to explore the stability and change of different types of victims and the key factors that influence their transition. Study 3 continues to track the same participants, and divides them into the constant population and the reconstructed population according to whether they change classes, to compare the differences of developmental patterns and predictive factors in victim subtypes between different population.

Participants were first-year students from seven secondary schools in a provincial capital city in central China. These seven schools include two urban schools, one country school, and four rural schools. A questionnaire survey was conducted at the end of each semester, starting from the first semester of seventh grade. This survey was conducted four times in total and collected 2,138 valid questionnaires. The average age of the participants at the first time (T1) was 12.99 years with a standard deviation of 0.62, and 49.5% were male. Victimization were obtained through both self- and peer-report methods. Depression and self-esteem were obtained through self-report methods. Popularity, the number of friends, and peer rejection were obtained through peer-report methods. The data analysis was conducted using SPSS 21.0 and Mplus 8.3.

The main results of the study are as follows:

(1) Based on the self- and peer-perceptions, there are three victim subtypes among teenagers, including self-identified victims (T1: 8.3%), peer-identified victims (T1: 7.2%), and non-victims (T1: 84.5%). The level of self-reported victimization is higher in self-identified victims, while the level of peer-reported victimization is lower. In contrast, the level of peer-reported victimization is higher in peer-identified victims, while the level of self-reported victimization is lower. The proportion of non-victims is higher than that of self-identified victims and peer-identified victims, indicating that most teenagers experience little or no victimization.

(2) There are some differences in the developmental patterns of different victim subtypes. The stability of non-victims is the highest (T1-T2: 92.8%), while the stability of self-identified victims (T1-T2: 64.8%) and peer-identified victims (T1-T2: 75.5%) is relatively low. Self-identified victims and peer-identified victims tend to transition to non-victims, indicating a gradual improvement in victimization.

(3) Internalizing problems and peer status have some predictive effects on the transition of different types of victims. Specifically, internalizing problems (depression and low self-esteem) are mainly related to the transition from self-identified victims to other subtypes of victims, such as depression leads to the transition from non-victims to self-identified victims, while low self-esteem hinders the transition from self-identified victims to non-victims. Peer rejection is mainly related to the transition of peer-identified victims to other subtypes of victims, such as hindering the transition of peer-identified victims to other subtypes of victims, and promoting other subtypes of victims to peer-identified victims.

(4) The stability of victim subtypes in the constant population (T3-T4) have changed, with the stability of self-identified victims (64.8% → 72.7%) and peer-identified victims (75.5% → 96.6%) increasing. Besides, the predictive effects of internalizing problems and peer status on the transition of different types of victims have decreased. Only peer rejection was found to promote the transition of non-victims to peer-identified victims.

(5) The stability and transition of victim subtypes in the reconstructed population (T3-T4) were relatively consistent with those in the initial stage of group formation (T1-T2). What is more, the predictive effects of internalizing problems and peer status on the transition of different types of victims have been restored. It not only found that peer rejection promoted the transition of non-victims to peer-identified victims and self-identified victims, but also found that depression to promoted the transition of non-victims to self-identified victims.

In summary, this study identifies three victim subtypes, including self-identified victims, peer-identified victims, and non-victims, while finding differences between them in the specific forms of victimization. Different types of victims have a certain cross-time stability, but there are still some individuals who transition to other subtypes of victims over time, with depression, self-esteem, and peer rejection playing a role in the transition process. In addition, the stability of victim subtypes in the constant population has increased, but the predictive effects of different transitional factors have weakened. While the stability of the victim subtypes and the predictive function have remained essentially the same in the reconstructed population.