Early Warning Systems for dropout prevention: Evidence from a randomized controlled trial in Guatemala

We administered:
a) A Brief manual/brochure providing general guidance on strategies to prevent dropout: The manual included suggestions for simple actions based on best international practices and a list of existing resources within the Ministry of Education to help with dropout, including scholarships and remedial tutoring programs and all necessary information to access these resources.
b) A list of at-risk students in each school. By applying linear regressions and basic prediction concepts to routinely collected administrative data, a prediction model identified the students in the 6th grade that were at high risk of dropping out in the transition to 7th grade and this list was provided to schools.
c) Reminders and social recognition. Periodic reminders were sent to school principals to remind them the importance of dropout prevention. In addition, the Ministry of education informed that it will formally recognize and publicize the top performing schools/principals in terms of reductions in dropout rates.
The main objective of the intervention is to reduce dropout rates in the transition from 6th to 7th grade, benefiting 6th grade students who would otherwise leave school.

2018-05-29

2019-04-07

1) Final outcomes:
1.1) Short-term: Drop-out rate in the transition from 6th grade to secondary education, Retention rate at 6th grade, Grades in language, math, and the rest of subjects
1.2) Medium term: Rate of transition to upper secondary and tertiary education, Survival rates in the educational system, School life expectancy, Over-age at school

2) Intermediate outcomes (Very short-term - principal's reaction to the intervention):
Degree of focalization of support on students who need more help, Perception of drop-out rates in the transition from primary to secondary, Perception of the returns to education, Tasks that are prioritized by school principals, Perception of the principal's influence on the dropout decision, Perception of the main influence on the dropout decision, Degree of importance given to the completion of secondary education, Principal's perception about students' belief on the usefulness of their education, Degree of knowledge of students' families and frequency of meetings with them.

Intermediate outcomes will be measured by comparing changes between the Baseline and the Follow-up questionnaires, while Final outcomes will be measured using Administrative data provided by all schools in the SIRE (Sistema Integrado de Registros Educativos).

Short-term: School choice (modality, quality, distance, peers, type of administration (public-private)), Self-reported reason for drop-out

These final outcomes will be measured using Administrative data provided by all schools in the SIRE (Sistema Integrado de Registros Educativos).

We randomly assigned 4000 public primary schools in Guatemala to: (a) a control group (with no intervention), (b) a treatment group receiving a training and a brief guiding manual; (c) a treatment group receiving the same as group (b) plus a list of at-risk students; (d) a treatment group receiving the same as group (c) plus reminders and social recognition to school principals that were successful in reducing dropout.

Randomization done in office by a computer

Schools

Yes

4,000 schools

77,208 students

Control: 1,000 schools; Treatment arm 1: 1,000 schools; Treatment arm 2: 1,000 schools; Treatment arm 3: 1,000 schools