Experimental Impacts of a Preschool Intervention in Chile on Children's Language Outcomes: Moderation by Student Absenteeism

Mary Catherine Arbour, Hirokazu Yoshikawa, John Willett, Christina Weiland, Catherine Snow, Susana Mendive, M. Clara Barata, Ernesto Treviño

Research output: Contribution to journalArticlepeer-review


Despite consensus that school absenteeism has negative consequences for children's life outcomes, until recently, little was known about the prevalence of absenteeism or its potential to moderate the impacts of school-based interventions. This study provides evidence from a randomized experiment of a preschool intervention involving 1,876 children in 64 schools in Chile that chronic absenteeism develops in preschool and is predicted by multiple risk factors for poor academic achievement. We find moreover that individual children's likelihood of absenteeism moderated the intervention's impact on children's language and literacy outcomes such that there were positive impacts of the intervention only for children with the lowest likelihood of absenteeism. Experimental evaluations of school-based interventions that do not take absenteeism into account may thus mask heterogeneous effects. In the context of policy pushes to expand early education and preschool access in the United States and globally, these moderation analyses may prove essential for appropriately interpreting the results of experimental studies of school-based interventions.

Original languageEnglish (US)
Pages (from-to)117-149
Number of pages33
JournalJournal of Research on Educational Effectiveness
StatePublished - Oct 3 2016


  • Chile
  • Latin America
  • attendance
  • chronic absenteeism
  • experiment
  • impact moderation
  • preschool education
  • regression-based subgroup analysis

ASJC Scopus subject areas

  • Education


Dive into the research topics of 'Experimental Impacts of a Preschool Intervention in Chile on Children's Language Outcomes: Moderation by Student Absenteeism'. Together they form a unique fingerprint.

Cite this