Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/299316 
Year of Publication: 
2023
Series/Report no.: 
Research Group on Human Capital - Working Paper Series No. 23-01
Publisher: 
Université du Québec à Montréal, École des sciences de la gestion (ESG UQAM), Groupe de recherche sur le capital humain (GRCH), Montréal
Abstract: 
This study estimates the average multivalued treatment effects (ATET), of preschool attendance measured in years, on students' international reading, math and science test z-scores in Grade 4. The causal treatment effects come from multiple-years observational data on three levels of preschool duration before entering Grade 1. Among European countries that participated in five international education surveys, PIRLS (2006, 2011, 2016) and TIMSS (2015, 2019), those renowned for having adopted early childhood education (ECE) programs starting at a young age, growing in intensity and improving the number of qualified child-care providers were selected. In addition to four Nordic countries (Denmark, Finland, Norway, Sweden), France, two Belgium jurisdictions (French, Flemish), and two participating Canadian provinces, Ontario and Québec, were retained. The approach exploits the repeated surveys and cross-national comparative international z-scores tests. The data sets besides their test scores provide unique information from a parent questionnaire on their education and occupation levels, literacy and numeracy preschool activities, on child preschool educational childcare span in years and two program types (for some years; before and after age 3). Four key findings can be identified from the data sets and estimations. First, there are large differences in the average scale score and percentiles deviation when converted into the z-score metric, for all categories of test scores across jurisdiction participants, and over time. Second, the estimates of the preschool treatment effects display rather heterogeneous impacts on z-scores with increasing significant and positive achievements over year surveys. Third, in general, preschool treatment effects are scattered in function of duration, programs types, and parental education. Four, results highlight stark gaps in scores related to parental education, socioeconomic statuses, and home learning resources for all year-samples. Evidence from a diversity of estimated gradients suggests established social inequalities in education achievement at ages 9-10 in Grade 4 could be difficult to reverse, even in cases where preschool education and care are implemented at a very young age in rich countries with very generous family policies.
Subjects: 
Preschool education years intensities
fourth graders reading, math and science tests scores
multivalued treatment effects
PIRLS (2006, 2011, 2016)
TIMSS (2015, 2019)
Denmark,Finland, Norway, Sweden, Belgium French, Belgium Flemish, France
Canadian provinces of Québec and Ontario
JEL: 
I2
Document Type: 
Working Paper

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.