Early identification of emotional and behavioral disorders is critical in ensuring that students receive the interventions and supports necessary for school success. While externalizing and internalizing behaviors often occur comorbidly, more research is needed to understand how different subtypes of these behaviors may manifest, especially in the early elementary school years. Further, as schools increase their emphasis on universal, evidence-based interventions as tools for preventing the development of later social-emotional and behavior challenges, it is important to understand how different groups of students respond to such interventions. In the present study, I explored the behavioral and emotional profiles of 470 second-grade students using latent profile analysis. I also examined how students transitioned between profiles over the course of one school year (i.e., fall to spring) within the context of a social-emotional intervention (Social Skills Improvement System-Classwide Intervention Program, SSIS-CIP; Elliott & Gresham, 2007). Additionally, I used multinomial logistic regression analyses to examine if child race, gender, teacher-student relationship, and treatment condition (intervention vs. control) predicted profile membership and transition over the course of a school year. I used five behavioral composites from the Social Skills Improvement System Rating Scales-Teacher Form (SSIS-RST; Gresham & Elliott, 2008) to create the student profiles. Three profiles of students emerged in the present study. The first profile (normative) was characterized by the lowest levels of all five externalizing and internalizing behaviors. The second profile (at-risk) demonstrated elevated levels of impulsive behaviors, conduct problems, and emotion dysregulation, with less elevated levels of bullying. The third profile (comorbid) demonstrated elevated levels of all five externalizing and internalizing behaviors. From fall to spring, students in the normative profile exhibited a 93% probability of remaining in the normative profile. Students in the at-risk profile demonstrated a similarly high likelihood of remaining in the at-risk profile over time (72%), while students in the comorbid profile experienced a 57% chance of remaining in the comorbid profile over time. Teachers’ observed emotional support, child race, child gender, and participation in the SSIS-CIP intervention were all predictors of profile movement, however, the associations varied across the different profiles. Results of the present study suggest the SSIS-CIP may function as a preventative tool for students identified within the normative group as well as an effective intervention for those students with the most severe behavioral presentations. Consistent with previous research, teachers rated males and students of color higher on measures of externalizing behaviors. Notably, while female students were less likely to be identified in the comorbid profile in the fall, they were more likely to remain in that profile over time compared to male students who were identified in the comorbid profile in the fall. Surprisingly, teacher emotional support was negatively associated with movement from the at-risk group to the normative group. Future research should continue to explore the ways in which externalizing and internalizing behaviors manifest in young children as well as the intersection between gender and race as it relates to teachers’ ratings of students’ behaviors. / School Psychology

L’efficacité des interventions pour diminuer la victimisation par les pairs en milieu scolaire est modeste. Un manque de connaissance sur le développement de la victimisation par les pairs en début de scolarité, notamment sur les rôles adoptés par les jeunes enfants (p. ex., agresseurs, victimes) ainsi que sur la contribution des divers comportements sociaux qu’ils exhibent, pourrait contribuer à cette efficacité limitée. Pour y remédier, il est nécessaire d’avoir une meilleure compréhension des enjeux propres à cette période développementale. Ainsi, cette thèse vise à étudier les différents profils de comportements sociaux et de victimisation par les pairs chez les enfants de la maternelle à la deuxième année du primaire, et ce, à l’aide d’un échantillon de 1757 enfants provenant de deux cohortes de naissance (Étude longitudinale sur le développement des enfants du Québec - ELDEQ). Plus spécifiquement, un premier article vise à explorer le développement de ces différents profils de la maternelle à la deuxième année, ainsi que leur stabilité d’une année à l’autre. Des analyses de profils latents sont effectuées à l’aide de mesures annuelles de l’agressivité physique, de l’agressivité relationnelle, de la victimisation subie et des comportements prosociaux des enfants, suivies par des analyses de transitions latentes. Les résultats montrent que les garçons présentent les quatre mêmes profils chaque année (prosociaux, normatifs, physiquement agressifs-victimisés et agressifs-victimisés) alors que les filles en présentent quatre en maternelle (prosociales, normatives, socialement agressives et agressives-victimisées) et cinq en première et deuxième année (prosociales, normatives, victimisées, socialement agressives, et agressives-victimisées). À la différence de la typologie classique, les garçons ne présentent pas de profil uniquement victimisé ou uniquement agressif et les profils des filles présentent des nuances sur la forme d’agressivité privilégiée. Une fois établie, l’appartenance aux profils s’avère relativement stable d’une année à l’autre, indiquant la précocité des problèmes sociaux avec les pairs. Afin de valider cette typologie des comportements sociaux et de la victimisation par les pairs et de suggérer des cibles potentielles d'évaluation ou d’intervention, le deuxième article teste d’abord si les habiletés cognitives préscolaires prédisent l’appartenance aux profils identifiés. Ensuite, il vérifie si l’expérience scolaire concomitante (relation enseignant-élève, engagement scolaire) diffère selon l’appartenance aux profils. Des analyses de régression multinomiale montrent que des habiletés cognitives élevées prédisent une plus grande probabilité de présenter un profil prosocial plutôt qu’agressif-victimisé. L’analyse des variables concomitantes montre que l’engagement scolaire est plus faible et la relation enseignant-élève plus difficile pour les enfants agressifs-victimisés, en comparaison aux enfants prosociaux ou aux filles socialement agressives en première et deuxième année. Ces résultats soulignent les associations précoces entre les difficultés sociales et scolaires, soit dès l'entrée à l'école. Pour finir, les résultats de cette thèse confirment l’existence et la pertinence d’une typologie des comportements sociaux et de la victimisation par les pairs en début de scolarité, soulignant la stabilité de celle-ci et la précocité des problèmes associés. Des cibles sont identifiées pour améliorer le dépistage des enfants à risque et, éventuellement, l‘efficacité des interventions préventives. Les différences garçons-filles, les caractéristiques uniques du profil des filles socialement agressives ainsi que l’expérience scolaire difficile des enfants agressifs-victimisés dès la maternelle sont discutées. / Interventions to reduce peer aggression and victimization in school settings currently show modest effectivity, lead on in part by limited knowledge on the development of peer victimization during the early school years. More specifically, there is a lack of information about which roles, or profiles, linked to peer aggression and victimization exist from school entry onwards, and how they change over time, as well as how children’s social behaviours are associated to them. To address this issue, this thesis aims to understand the different profiles of social behaviour and peer victimization that children can present from kindergarten to second grade. The study included 1757 children from two birth cohorts in the Quebec (Quebec Longitudinal Study of Child Development - QLSCD). More specifically, the first study included in this thesis explores the development of different profiles related to social behaviour and peer victimization from kindergarten to second grade, using annual measures of children's physical aggression, relational aggression, victimization, and prosocial behaviours. The results of the latent profile analyses performed show that boys present the same four profiles each year: prosocial, normative, physically aggressive-victimized (physically AV) and aggressive-victimized (AV). Girls present four profiles in kindergarten (prosocial, normative, socially aggressive and AV) and five in first and second grade (prosocial, normative, victimized, socially aggressive, and AV). Latent transition analyses aimed at testing the stability of profile membership over time showed that children tend to stay in the same profile over the years, especially if they are prosocial or AV. This study offers a slightly different typology from the one that is typically used in peer victimization theories, and it highlights the precocity of social behaviour and peer victimization problems for some children, stressing the importance of early identification and intervention. The second study of this thesis validates the typology of social behaviour and peer victimization and aims to identify potential targets for evaluation and intervention. It tests whether preschool cognitive abilities predict profile membership from kindergarten to second grade and tests whether children’s concurrent school experience differs according to their profile. For the first aim, results showed that high cognitive abilities predict a greater likelihood of exhibiting a prosocial rather than an AV role when controlling for socioeconomic status. For 6 the second aim, results showed that the quality of AV children’s school engagement and relationship with their teacher is worse than those of prosocial children at all time points, or of socially aggressive girls in first and second grade. Once again, these differences highlighted how early difficulties associated with children’s social problems can start. In conclusion, this thesis confirms the existence and relevance of a typology of social behaviour and peer victimization at the start of schooling, highlighting the stability of profile membership and the precocity of associated problems. Targets are identified to improve screening and the effectiveness of preventive interventions. Differences between boys’ and girls’ typologies and school experience, the uniqueness of the socially aggressive girl profile, and the struggles of AV children from kindergarten onwards are discussed.