Although the concept of school engagement figures prominently in most school dropout theories, there has been little empirical research conducted on its nature and course and, more importantly, the association with dropout. Information on the natural development of school engagement would greatly benefit those interested in preventing student alienation during adolescence. Using a longitudinal sample of 11,827 French‐Canadian high school students, we tested behavioral, affective, cognitive indices of engagement both separately and as a global construct. We then assessed their contribution as prospective predictors of school dropout using factor analysis and structural equation modeling. Global engagement reliably predicted school dropout. Among its three specific dimensions, only behavioral engagement made a significant contribution in the prediction equation. Our findings confirm the robustness of the overall multidimensional construct of school engagement, which reflects both cognitive and psychosocial characteristics, and underscore the importance attributed to basic participation and compliance issues in reliably estimating risk of not completing basic schooling during adolescence.

Although most theories draw upon the construct of school engagement in their conceptualization of the dropout process, research addressing its hypothesized prospective relation with dropout remains scarce and does not account for the academic and social heterogeneity of students who leave school prematurely. This study explores the reality of different life-course pathways of school engagement and their predictive relations to dropout. Using an accelerated longitudinal design, we used growth mixture modeling to generate seven distinct trajectories of school engagement with 12-to 16-year-old students (N = 13,300). A vast majority of students were classified into three stable trajectories, distinguishing themselves at moderate to very high levels of school engagement. We refer to these as developmentally normative pathways in light of their frequent occurrence and stability. Although regrouping only one-tenth of participants, four other nonnormative (or unexpected pathways) accounted for the vast majority of dropouts. Dropout risk was closely linked with unstable pathways of school engagement. We conclude by debating the delicate investment balance between universal strategies and more selective and differentiated strategies to prevent dropout. We also discuss the need to better understand why, within normative trajectories, some students with high levels of school engagement drop out of school.From a production of human capabilities perspective, which shares its preoccupations across disciplines concerned with children's development (Heckman, 2006), our industrial era is as exciting as it is daunting. The constant evolution of knowledge and its concomitant technology poses great demands on learners and educational systems alike (OECD, 2006). Consequently, governments have become increasingly concerned with performance indicators and generating policies that will increase the quality of both academic and social learning, especially for culturally and economically disadvantaged populations (Heckman, 2006;OECD, 2006). The goal is successful output of qualified students at the most basic level

School-based interventions should address the multiple facets of high school experiences to help adolescents successfully complete their basic schooling. Creating a positive social-emotional learning environment promises better adolescent achievement and, in turn, will contribute to a healthier lifestyle.

We first replicated the data analytic strategy used in Duncan et al. (2007) with a population-based data set of French-speaking children from Quebec (Canada). Prospective associations were examined between cognitive, attention, and socioemotional characteristics underlying kindergarten school readiness and second grade math, reading, and general achievement. We then extended this school readiness model by including motor skills as an additional element in the prediction equation and expanded the original strategy by including classroom engagement. The Montreal Longitudinal-Experimental Preschool Study, featured in Duncan et al., served as the Canadian reference group. In the replication model, kindergarten cognitive and attention characteristics predicted achievement by the end of 2nd grade. Although inconsistent across outcomes, behavioral problems and skills also emerged as predictors of some aspects of later achievement. Coefficients for kindergarten math skills were largest, followed by attention skills, receptive language skills, attention problems, and behavior. Most coefficients resembled those generated in the initial study. In our extension model, fine motor skills added their significant contribution to the prediction of later achievement above and beyond the original key elements of school readiness. Our extension model confirmed prospectively associations between kindergarten cognitive, attention, fine motor, and physical aggression characteristics and later achievement and classroom engagement by the end of 2nd grade. Although they comparatively showed better long-term benefits from stronger early attention skills, girls with less kindergarten cognitive skills were more vulnerable than boys with similar deficits when predicting 2nd grade math.

Because literacy skills are critical for most academic subject matters, researchers have become increasingly interested in understanding children's motivation in this domain as a way to increase academic success. In this study, we extend previous work by looking at the heterogeneity of children's motivational changes in literacy across Grades 1-12. We used a cross-sequential design based on 3 different cohorts of children (N ϭ 655) coming from 10 public elementary schools. Data were collected over an 8-year period (1988 -1996) starting when children were in 1st, 2nd, and 4th grades. Using multiple-process growth mixture modeling, we identified 7 groups of children showing distinct trajectories of change in literacy subjective task value and ability self-concept across school years. For all children, ability self-concept and subjective task value decreased between Grade 1 and Grade 12; however, this decline was more apparent for some youths than for others. Our findings suggest the importance of school-based interventions that aim to prevent decreases in student motivation for literacy over time and that focus on the early detection of children at risk for dramatic declines.