Back to search

FINNUT-Forskning og innovasjon i utdanningssektoren

Transgenerational, Social, and Individual Predictors of High-School Dropout: An Ecological Model Tested in a Multi-Data Design

Alternative title: Distale og proksimale faktorer for frafall i videregående skole - en holistisk modell undersøkt gjennom flerstudieanalyser.

Awarded: NOK 12.3 mill.

More than 25% of Norwegian students in upper secondary school (16-19 years old) do not graduate within five years of enrollment. Much indicates that dropout from upper secondary school is a result of a process that begins early in life, and that youth who drop out of secondary education are vulnerable to a range of social and health problems. There is fairly widespread agreement that interventions preventing factors that amplify vulnerability to dropout in an early phase are more successful. In DROPOUT, we identify the effects of generational, social, and individual factors, as well as map the complex interaction between such factors from childhood to adolescence using a multi-data design and innovative statistical methods. We benefit from data from informants with up to eight measurement points (Early Safe in Trondheim study), along with data spanning multiple generations with rich opportunities for linkages to registry data (The Health Survey in Trøndelag). Analyses of this material have commenced, in collaboration with national and international partners. Preliminary findings in the project show, among other things, that mothers' mental health negatively influences the child's school performance between the ages of 8 and 14, and more specifically, that mothers' anxiety and depression negatively affect to an equal extent. In the same study, we also found that the child's lack of school performance reinforced the mother's depressive symptoms. We have also found that the teacher-student relationship is of great importance for students' academic performance in the transition to upper secondary school but has little impact before this. The relationship with mother and father is more significant in the early school years. We have also found that socially withdrawn boys become less withdrawn through academic achievement, which does not apply to girls. Previously, it was primarily thought that children experienced reduced school performance due to shyness, but this was not found; only that an improvement in academic mastery can prevent social withdrawal, but only among boys. We have also found that internal motivation matters more than cognitive abilities for improvement in school performance in secondary school, indicating that the school context and the facilitation of an autonomous learning environment are crucial for academic achievement. Furthermore, we examine, for example, how social interaction with peers in the primary and secondary school years can threaten the conditions for good learning in school by weakening the ability to self-regulate, something we have repeatedly found to occur when children and youth experience social exclusion. These are some of the tracks we follow that will eventually be significant for dropout or opting out of upper secondary school. Data to use dropout in upper secondary as the outcome variable in The Early Seccure in Trondheim study will be ready for analysis in the spring of 2024.

More than 25% of Norwegian high-school students (16-19 years) fail to graduate within five years after admission (n>10 000). Youth dropping out from school are at risk for a multitude of problems, and the current COVID-19 pandemic is likely to excel the number of individuals vulnerable to social marginalization. High-school dropout is often just an endpoint of a process that for many starts early in life. Hence, policies and interventions preventing dropout trajectories at an early stage bear promise of greater success than remediating efforts. In DROPOUT, we propose to identify the impact of generational, social, and individual factors, as well as map the complex interplay of such factors from childhood to adolescence by employing a multi-data design and innovative statistical methods. Even prospective studies are at risk for confusing correlate with cause. Hence, recent statistical approaches, focusing on within-person analyses, reduce this risk of false conclusions, by discounting all unmeasured time-invariant confounders—including genetics. Moreover, within-person analyses may be expanded to also adjust for measured time-varying confounding, thereby further reducing the risk of mischaracterizing a correlate as a causal factor. We will take advantage of prospective multi-informant data with up to 8 measure points (The Trondheim Early Secure Study) together with data spanning over several generations with rich possibilities for links to register data (The Trøndelag Health Study). These comprehensive data pools, containing different assets with respect to variable complexity, measure point proximities, and representativeness/statistical power, combined with the team's high statistical competence, put our research team in position to provide important insights into the short-term and the long-term causal relationships leading to increased/decreased likelihood for high-school dropout.

Funding scheme:

FINNUT-Forskning og innovasjon i utdanningssektoren