The International Observatory on Academic Achievement - Documentary search

File number :
CS-DSC-12e

Bibliographic reference :
Janosz, M., LeBlanc, M., Boulerice, B., & Tremblay, R. E. (1997).  Disentangling the Weight of School Dropout Predictors:  A Test on Two Longitudinal Samples. Journal of Youth and Adolescence, 26(6), 733-762.

Abstract :

Theoretical Context/Study Objectives
The authors of this Quebec study (Canada) began by presenting the main predictors of school dropout. Based on studies conducted over the past 25 years, they identified the key individual, family and school risk factors associated with school dropout. However, they specified that each factor has a different weight in understanding the problem. They underscored the importance of adequately identifying the different variables for explaining school dropout in order to implement effective screening and prevention programs.

The authors of this study targeted two main objectives: i) identify factors that predict dropout as precisely as possible, despite differences among subjects and over time; ii) then, attempt to determine the stability of the most general, predictive factors among subjects and over time.

Methodology
Two independent, longitudinal samples, each comprised of 791 subjects were used in this study. The first, representative of the Montréal region, was part of a study on delinquency begun in 1974. The second sample was from a psychosocial adjustment study of teenagers with low to moderate socio-economic status (SES). In 1985, at least one year prior to leaving school, these students completed a questionnaire to gather personal (social adjustment, aggressiveness, anxiety and so on) and social data (school and family experience, interpersonal relationships with peers, leisure pursuits, commitment to norms, deviant behaviour).

Results
In the first sample, 22% of young people were identified as dropouts, while 42% of young people in the second sample had quit school. The authors affirmed that 74% of predictive factors found in the 1974 study were also present in the 1985 study, while 81% of factors identified in 1985 were also present in 1974. Therefore, a majority of risk factors (school, family, behavioural, social and psychological) proved stable over time and across samples.

Next, with respect to the relative importance of predictors, the authors demonstrated that school factors are the most powerful predictors of school dropout, even more so among high-risk populations. Specifically, these factors are grade repetition, academic performance and school engagement. A brief questionnaire verifying these variables could suffice to determine which students are at risk.

For more details, see file CS-DSC-13.

Links :
This journal is also available in electronic format.

Key Words :
Developmental perspective, Predictive factors/Predictors, Personal characteristics, Family characteristics, School characteristics, Grade repetition, School engagement, Longitudinal study, Quantitative analyse analysis, Secondary/High school

Monitored Countries :
Quebec (Canada)